lint: Removed accidental session file

The Client trait's default chat_completions, chat_completions_streaming,
and embeddings methods now classify failures via ApiStatusError: on a
401 with a cached OAuth token, the token is distrusted (identity-aware
marker) and the call retried exactly once — the retry's prepare step
sees the marker and force-refreshes. Streaming retries only while the
SSE handler has received no content, preventing duplicate rendering.
A second 401 propagates the original error; other retry errors
propagate as-is. API-key clients never retry. No backoff by design:
cost is bounded to one refresh + one retry per failing request.

.github/ISSUE_TEMPLATE/bug_report.yaml

@@ -21,25 +21,25 @@ body:
       value: |
         I tried this:
 
-        1. `loki`
+        1. `coyote`
 
         I expected this to happen:
 
         Instead, this happened:
   - type: textarea
-    id: loki-log
+    id: coyote-log
     attributes:
-      label: Loki log
-      description: Include the Loki log file to help diagnose the issue. (`loki --info` to see the log_path)
+      label: Coyote log
+      description: Include the Coyote log file to help diagnose the issue. (`coyote --info` to see the log_path)
       value: |
         | OS      | Log file location                                     |
         | ------- | ----------------------------------------------------- |
-        | Linux   | `~/.cache/loki/loki.log`                      |
-        | Mac     | `~/Library/Logs/loki/loki.log`                |
-        | Windows | `C:\Users\<User>\AppData\Local\loki\loki.log` |
+        | Linux   | `~/.cache/coyote/coyote.log`                      |
+        | Mac     | `~/Library/Logs/coyote/coyote.log`                |
+        | Windows | `C:\Users\<User>\AppData\Local\coyote\coyote.log` |
 
         ```
-        please provide a copy of your loki log file here if possible; you may need to redact some of the lines
+        please provide a copy of your coyote log file here if possible; you may need to redact some of the lines
         ```
 
   - type: input
@@ -57,13 +57,13 @@ body:
     validations:
       required: true
   - type: input
-    id: loki-version
+    id: coyote-version
     attributes:
-      label: Loki Version
+      label: Coyote Version
       description: >
-        Loki version (`loki --version` if using a release, `git describe` if building
+        Coyote version (`coyote --version` if using a release, `git describe` if building
         from main).
-        **Make sure that you are using the [latest loki release](https://github.com/Dark-Alex-17/loki/releases) or a newer main build**
-      placeholder: "loki 0.1.0"
+        **Make sure that you are using the [latest coyote release](https://github.com/Dark-Alex-17/coyote/releases) or a newer main build**
+      placeholder: "coyote 0.1.0"
     validations:
       required: true

.github/workflows/ci.yaml

@@ -36,6 +36,17 @@ jobs:
 
     - uses: Swatinem/rust-cache@v2
 
+    - name: Cache DuckDB Extensions
+      id: duckdb-extensions
+      uses: actions/cache@v4
+      with:
+        path: ~/.duckdb/extensions
+        key: duckdb-ext-${{ matrix.os }}-${{ hashFiles('Cargo.lock') }}
+
+    - name: Install DuckDB Extensions
+      if: steps.duckdb-extensions.outputs.cache-hit != 'true'
+      run: cargo test --all duckdb
+
     - name: Test
       run: cargo test --all
 

.github/workflows/release.yaml

@@ -8,9 +8,9 @@ on:
   workflow_dispatch:
     inputs:
       bump_type:
-        description: "Specify the type of version bump"
+        description: 'Specify the type of version bump'
         required: true
-        default: "patch"
+        default: 'patch'
         type: choice
         options:
           - patch
@@ -46,7 +46,7 @@ jobs:
       - name: Set up Python
         uses: actions/setup-python@v4
         with:
-          python-version: "3.10"
+          python-version: '3.10'
 
       - name: Install Commitizen
         run: |
@@ -98,9 +98,9 @@ jobs:
           # Ignore Act's local artifact dir noise
           echo artifacts/ >> .git/info/exclude || true
 
-          # Edit the version line right after name="loki"
+          # Edit the version line right after name="coyote"
           sed -E -i '
-            /^[[:space:]]*name[[:space:]]*=[[:space:]]*"loki"[[:space:]]*$/ {
+            /^[[:space:]]*name[[:space:]]*=[[:space:]]*"coyote"[[:space:]]*$/ {
               n
               s|^[[:space:]]*version[[:space:]]*=[[:space:]]*"[^"]*"|version = "'"$VERSION"'"|
             }
@@ -108,17 +108,19 @@ jobs:
 
           cargo update || true
 
+          sed -i "s|image: 'darkalex17/coyote:v[^']*'|image: 'darkalex17/coyote:v${VERSION}'|" assets/sbx-kit/spec.yaml
+
           # Git config that helps in Act
           git config user.name  "github-actions[bot]"
           git config user.email "github-actions[bot]@users.noreply.github.com"
           git config --global --add safe.directory "$GITHUB_WORKSPACE"
 
           git status --porcelain
-          git diff --name-only -- Cargo.toml Cargo.lock || true
+          git diff --name-only -- Cargo.toml Cargo.lock assets/sbx-kit/spec.yaml || true
 
-          if ! git diff --quiet -- Cargo.toml Cargo.lock; then
-            git add -u -- Cargo.toml Cargo.lock
-            git commit -m "chore: bump Cargo.toml to $VERSION"
+          if ! git diff --quiet -- Cargo.toml Cargo.lock assets/sbx-kit/spec.yaml; then
+            git add -u -- Cargo.toml Cargo.lock assets/sbx-kit/spec.yaml
+            git commit -m "chore: bump Cargo.toml and sandbox image to $VERSION"
           else
             echo "No changes to commit (already at $VERSION)"
           fi
@@ -163,28 +165,28 @@ jobs:
           - target: aarch64-unknown-linux-musl
             os: ubuntu-latest
             use-cross: true
-            cargo-flags: ""
+            cargo-flags: ''
           - target: aarch64-apple-darwin
             os: macos-latest
             use-cross: true
-            cargo-flags: ""
+            cargo-flags: ''
           - target: aarch64-pc-windows-msvc
             os: windows-latest
             use-cross: true
-            cargo-flags: ""
+            cargo-flags: ''
           - target: x86_64-apple-darwin
             os: macos-latest
-            cargo-flags: ""
+            cargo-flags: ''
           - target: x86_64-pc-windows-msvc
             os: windows-latest
-            cargo-flags: ""
+            cargo-flags: ''
           - target: x86_64-unknown-linux-musl
             os: ubuntu-latest
             use-cross: true
-            cargo-flags: ""
+            cargo-flags: ''
           - target: x86_64-unknown-linux-gnu
             os: ubuntu-latest
-            cargo-flags: ""
+            cargo-flags: ''
 
     steps:
       - name: Check if actor is repository owner
@@ -278,7 +280,7 @@ jobs:
       - name: Verify file
         shell: bash
         run: |
-          file target/${{ matrix.target }}/release/loki
+          file target/${{ matrix.target }}/release/coyote
 
       - name: Test
         if: matrix.target != 'aarch64-apple-darwin' && matrix.target != 'aarch64-pc-windows-msvc'
@@ -338,7 +340,7 @@ jobs:
             ${{ steps.package.outputs.archive }} 
             ${{ steps.package.outputs.sha }}
           tag_name: v${{ env.RELEASE_VERSION }}
-          name: "v${{ env.RELEASE_VERSION }}"
+          name: 'v${{ env.RELEASE_VERSION }}'
           body_path: artifacts/changelog.md
           prerelease: false
 
@@ -382,11 +384,11 @@ jobs:
         shell: bash
         run: |
           # Set environment variables
-          macos_sha="$(cat ./artifacts/loki-x86_64-apple-darwin.sha256 | awk '{print $1}')"
+          macos_sha="$(cat ./artifacts/coyote-x86_64-apple-darwin.sha256 | awk '{print $1}')"
           echo "MACOS_SHA=$macos_sha" >> $GITHUB_ENV
-          macos_sha_arm="$(cat ./artifacts/loki-aarch64-apple-darwin.sha256 | awk '{print $1}')"
+          macos_sha_arm="$(cat ./artifacts/coyote-aarch64-apple-darwin.sha256 | awk '{print $1}')"
           echo "MACOS_SHA_ARM=$macos_sha_arm" >> $GITHUB_ENV
-          linux_sha="$(cat ./artifacts/loki-x86_64-unknown-linux-musl.sha256 | awk '{print $1}')"
+          linux_sha="$(cat ./artifacts/coyote-x86_64-unknown-linux-musl.sha256 | awk '{print $1}')"
           echo "LINUX_SHA=$linux_sha" >> $GITHUB_ENV
           release_version="$(cat ./artifacts/release-version)"
           echo "RELEASE_VERSION=$release_version" >> $GITHUB_ENV
@@ -402,23 +404,23 @@ jobs:
         if: env.ACT != 'true'
         run: |
           # run packaging script
-          python "./deployment/homebrew/packager.py" ${{ env.RELEASE_VERSION }} "./deployment/homebrew/loki.rb.template" "./loki.rb" ${{ env.MACOS_SHA }} ${{ env.MACOS_SHA_ARM }} ${{ env.LINUX_SHA }}
+          python "./deployment/homebrew/packager.py" ${{ env.RELEASE_VERSION }} "./deployment/homebrew/coyote.rb.template" "./coyote.rb" ${{ env.MACOS_SHA }} ${{ env.MACOS_SHA_ARM }} ${{ env.LINUX_SHA }}
 
       - name: Push changes to Homebrew tap
         if: env.ACT != 'true'
         env:
-          TOKEN: ${{ secrets.LOKI_GITHUB_TOKEN }}
+          TOKEN: ${{ secrets.COYOTE_GITHUB_TOKEN }}
         run: |
           # push to Git
           git config --global user.name "Dark-Alex-17"
           git config --global user.email "alex.j.tusa@gmail.com"
-          git clone https://Dark-Alex-17:${{ secrets.LOKI_GITHUB_TOKEN }}@github.com/Dark-Alex-17/homebrew-loki.git
-          rm homebrew-loki/Formula/loki.rb
-          cp loki.rb homebrew-loki/Formula
-          cd homebrew-loki
+          git clone https://Dark-Alex-17:${{ secrets.COYOTE_GITHUB_TOKEN }}@github.com/Dark-Alex-17/homebrew-coyote.git
+          rm homebrew-coyote/Formula/coyote.rb
+          cp coyote.rb homebrew-coyote/Formula
+          cd homebrew-coyote
           git add .
-          git diff-index --quiet HEAD || git commit -am "Update formula for Loki release ${{ env.RELEASE_VERSION }}"
-          git push https://$TOKEN@github.com/Dark-Alex-17/homebrew-loki.git
+          git diff-index --quiet HEAD || git commit -am "Update formula for Coyote release ${{ env.RELEASE_VERSION }}"
+          git push https://$TOKEN@github.com/Dark-Alex-17/homebrew-coyote.git
 
   publish-crate:
     needs: publish-github-release
@@ -455,4 +457,64 @@ jobs:
       - uses: katyo/publish-crates@v2
         if: env.ACT != 'true'
         with:
-          registry-token: ${{ secrets.CARGO_REGISTRY_TOKEN }}
+          registry-token: ${{ secrets.CARGO_REGISTRY_TOKEN }}
+
+  publish-sandbox-image:
+    needs: [publish-github-release]
+    name: Publish Sandbox Docker Image
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check if actor is repository owner
+        if: ${{ github.actor != github.repository_owner && env.ACT != 'true' }}
+        run: |
+          echo "You are not authorized to run this workflow."
+          exit 1
+
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 1
+
+      - name: Ensure repository is up-to-date
+        if: env.ACT != 'true'
+        run: |
+          git fetch --all
+          git pull
+
+      - name: Get release artifacts
+        uses: actions/download-artifact@v4
+        with:
+          path: artifacts
+          merge-multiple: true
+
+      - name: Set version variable
+        run: |
+          version="$(cat artifacts/release-version)"
+          echo "version=$version" >> $GITHUB_ENV
+
+      - name: Validate release environment variables
+        run: |
+          echo "Release version: ${{ env.version }}"
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        if: env.ACT != 'true'
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKER_USERNAME }}
+          password: ${{ secrets.DOCKER_PASSWORD }}
+
+      - name: Push to Docker Hub
+        uses: docker/build-push-action@v5
+        with:
+          context: .
+          file: Dockerfile
+          platforms: linux/amd64,linux/arm64
+          push: ${{ env.ACT != 'true' }}
+          tags: darkalex17/coyote:latest, darkalex17/coyote:v${{ env.version }}
+          build-args: COYOTE_VERSION=${{ env.version }}

.gitignore

@@ -3,5 +3,9 @@
 /.env
 !cli/**
 .idea/
-/loki.iml
+/coyote.iml
 /.idea/
+.coyote/**
+.sisyphus/**
+.coyote-project.json
+.coyote/memory/

.loki-project.json

@@ -1 +0,0 @@
-{"type":"rust","build":"cargo build","test":"cargo test","check":"cargo check","_detected_by":"heuristic","_cached_at":"2026-04-13T13:36:33-06:00"}

CHANGELOG.md

@@ -1,3 +1,468 @@
+## v0.8.3 (2026-08-03)
+
+### Fix
+
+- infinite loop bug when attempting to interrupt a prompt exchange right before a session compression
+- ctrl-c inside of an auto-continue loop created an infinite loop
+
+## v0.8.2 (2026-07-31)
+
+### Fix
+
+- sbx update doesn't allow undefined fields in sbx spec
+
+## v0.8.1 (2026-07-30)
+
+### Feat
+
+- ctrl-c interrupts ongoing prompt in a session, but lets the user inject more instructions mid-stream
+- improved function calling performance by allowing parallel tool calling
+- created the architect and gatekeeper agents for dramatically improved coding performance
+- Improved readability of session message exchange replays
+- apply --agent/--role/--rag/--model flags in --acp-server mode
+- add headless profile to sbx-kit spec
+- implement ACP user-interaction to request_permission bridge
+- implement ACP session/load and session/cancel
+- implement ACP session/prompt
+- add ACP server skeleton with stdout-purity test
+- add --headless flag for unattended operation
+
+### Fix
+
+- ctrl-c interruption doesn't discard session messages when throbber is showing
+- improper handling of fd-style globbing for directories in fs_glob
+- properly templated architect design doc path in starter commands
+- .copy works when sessions are resumed
+- ACP session/prompt now drives the full tool-execution loop
+- ACP spec conformance — ContentBlock prompt params and protocolVersion type
+- restore stdout output for standalone --headless mode
+- suppress tool-call display in headless mode; initialize session on session/new
+- skip stdin drain and set silent render mode when --acp-server is active
+- include graph-agent descriptions in .agent <TAB> completions
+
+### Refactor
+
+- move ACP server dispatch into run() for shared flag setup
+
+## v0.8.0 (2026-07-25)
+
+### Feat
+
+- Dynamically detect if a selected client in the sandbox first run wizard supports oauth
+- Add support for the --fresh flag again with host environment configuration injection
+- force overwrite global sbx secrets
+- add sbx secrets globally
+- only create secrets local to a sandbox
+- Improved credentials management for docker sandboxes
+- renamed --contents arg for fs_write/patch to --content since most models attempt that first and error otherwise
+- Used improved theme selections for tool call highlighting colors
+- Improved theme-derived syntax highlighting for LLM tool call logging
+- Improved coloring of LLM tool invocation outputs to make LLM output more readable and cohesive
+- renamed the user__ask to user__select and improved descriptions to improve model usage
+- Improved coloring/highlighting of tool calls to make LLM invocation logs easier to read
+- long-running session improvements
+- Made sisyphus suite of agents all auto-approved for tool usage
+- added ast_grep tool to Sisyphus suite agents
+- created the adversay agent and adversarial-review skill
+- new spawnable_agents field in agents to let users restrict what agents can be spawned by a parent agent
+- replay pre-compressed messages as well when resuming sessions for users to see
+- created a new builtin function for agents who can spawn other agents to list available agents via agent__list_available
+- **render**: hanging-indent line wrapping for lists and blockquotes
+- **render**: wire table state machine and finalize hook
+- **render**: render markdown tables with comfy-table
+- **render**: parse table cells and column alignments
+- **render**: detect markdown table rows and separators
+- **render**: add comfy-table dependency and table border style
+- **render**: activate rich markdown renderer as default
+- **render**: rich block-level markdown rendering (headings, quotes, lists, hr)
+- **render**: rich inline markdown rendering (bold, italic, code, links)
+- **render**: detect markdown block-level line types
+- **render**: precompute markdown scope styles for rich rendering
+- Created the raw_markdown configuration flag
+- added a .fork command to fork a new session from a running conversation
+- **oauth**: enable browser-paste PKCE flow for OpenAI-compatible providers
+- copy host OAuth tokens into sandbox at launch
+- implement OAuth 2.0 Device Authorization Grant (RFC 8628)
+- hint that browser 'paste code' pages can be ignored during callback capture
+- openai-compatible wizard offers OAuth when provider has bundled oauth defaults
+- validate unique client names at config load
+- bundle xAI OAuth defaults in models.yaml
+- OAuth branch in openai_compatible prepare_* fns
+- get_oauth_provider_for_client dispatcher + client_config_info update
+- OpenAICompatibleOAuthProvider (config-driven OAuthProvider impl)
+- add auth + oauth fields to OpenAICompatibleConfig
+- add oauth field to ProviderModels
+- add client_credentials support to prepare_oauth_access_token
+- add OAuthConfig + OAuthFlow types to oauth.rs
+- Added reasoning effort to the right prompt
+- Improved support for Anthropic's extended thinking
+- Also support GEMINI.md workspace instructions
+- Improved workspace instructions support
+- also detect .mcp.json configurations at workspace roots
+- Created new .tool enable/disable and .mcp enable/disable aliases to make REPL usage more egonomic
+- Created a new .list <kind> REPL command to make discoveribility easier in the REPL
+- Support claude-style hidden workspace MCP configuration files via .mcp.json
+- Allow users to customize the workspace-specific configuration directory name so they can use Coyote with other CLI clients like .claude
+- Add reasoning_effort validation for the main configuration file
+- Add validation for reasoning_effort settings to prevent users from specifying erroneous values
+- Added support for modifying the reasoning effort of reasoning models
+- Explicitly Prevent .undo usage in graph agents
+- Added an .undo command to the REPL to let users have more control over the conversation
+- Improve sandbox startup time by using the prebuilt Coyote image
+- Make coyote available as a docker image
+- Made fs_patch more flexible for different model preferences of patch formats
+- Installed nano into the sandbox so that users can edit config files in the sandbox directly
+- Support workspace-local skill definitions and MCP configurations
+- fully functional graph-based RAG
+- Implemented graph-based RAG
+- Added a --dangerously-skip-permissions flag to skip permission prompts for tool invocations
+- Remove the temperature hyperparameter from the diagnose role
+- Added a new oauth.redirectHost field to make it possible to further extend MCP support
+- Updated the REPL mcp auth path to use the prettified error messaging
+- Improved error messaging for failed MCP starts because of auth issues
+- Added support for specifying the oauth port and client ID in MCP server configs
+- Implemented OAuth support for OpenAI models via Codex endpoints
+- merge MCP config when installing bundled mcp config
+- Implemented durable state for sisyphus
+- Installed ast-grep for the explore agent to use for better code exploration
+- Created the step-runner graph agent for more deterministic coding workflows to produce even more reliable and higher-quality results
+- Improved oracle and sisyphus agents with skill integrations for the new skills
+- Created new sisyphus family skills to improve performance
+- Created new diagnostic role and skill for use in other contexts
+- Added new memory functions for deleting and renaming memory files, as well as new lints for memory expiration dates and staleness of memories to improve the memory system
+- Created a new iwe skill and installed the iwe MCP server for utilizing large knowledgebases
+- Session-specific, file-backed history in the REPL
+- Replay session output when a user re-enters a session so all output can be seen again
+- Added confirmation message after MCP Oauth succeeds when invoked from --auth-mcp
+- Created the --auth-mcp CLI flag to let users auth with remote MCP servers without needing to be in the REPL
+- add OAuth authentication support for remote MCP servers
+- Added mixin for sisyphus so the ddg MCP server can search arbitrary domains
+- added improved error messaging on MCP server initialization
+- prefer musl versions for linux when running --update/.update
+
+### Fix
+
+- fresh wizard openai-compatible support
+- config existence check for --fresh sandboxes
+- Improve coyote sandbox startup time
+- bypass forgotten sandbox mode check for MCP secret interpolation
+- properly wrap sub-style changes in markdown rendering
+- removed accidental duplicate ast_grep tool in explore agent
+- npm and npx need the /usr/local/share/npm-global/lib directory to exist to run properly so I've added it to the dockerfile
+- added executable bit to adversary agent tools script
+- fetch descriptions from graph agent configs as well when listing agents
+- **render**: suppress blank lines above rendered table
+- chown the full sandbox cache dir, not just the coyote subdir
+- chown the whole coyote cache dir not just the oauth dir in the sandbox
+- fix typo in Gemini's generation_config property to use camelCase exclusively
+- **oauth**: treat missing expires_in as non-expiring device_code token
+- **oauth**: send Accept: application/json in device flow requests
+- OAuth callback listener skips speculative/malformed browser connections
+- resolve reasoning effort for the prompt for global defaults as well
+- Account for default model reasoning_effort when supplying that value for the REPL prompts
+- model narration included in history and between tool calls to prevent repetition
+- Don't terminate agent loops early for null tool output
+- reduce code duplication by reusing the new concrete_tool_names function in .list tools
+- Agent tools can only be modified via .tool enable/disable using tools in the allowed whitelist in the agent
+- re-render agent sessions when entering agents with either pre-configured agent_session or when entering an agent directly into a session
+- Per RFC 9728, enable dynamic discovery of OAuth endpoints in MCP using path-aware discovery
+- hot-attach to MCP servers that require auth after running .mcp auth <name>
+- Correctly inherit graph-global model for extractor model if none is defined
+- no cursor timeout when user scrolls away from ongoing streaming output
+- default to the nano or notepad when a configured editor is not found
+- When EDITOR, VISUAL, or config.editor is defined, don't verify via which
+- Added a loop exit condition for the diagnostics skill
+- Added directness clause to the diganose role to improve prompt
+- fs tools now output better error handling to guide the model more effectively
+- Make fs_read more tolerant of various arg invocation formats.
+- todo functions are injected properly to roles when roles have auto_continue: true and the REPL is started directly into the role
+- updated the redirect URI for OAuth MCP to use localhost since that's what is whitelisted, not 127.0.0.1
+- allow MCP OAuth refresh_token to be absent from initial token exchanges
+- Overrode the default JSON content-type for MCP OAuth so its properly application/x-www-form-urlencoded
+- typo in mcp file name
+- Added uvx wrapper for macos-based sandboxes
+
+### Refactor
+
+- Standardized paths module function names to not use 'path' in the name and to just always be either 'dir' or 'file'
+- main.rs resolve_oauth_client uses new dispatcher
+- split run_oauth_flow into pkce + client_credentials dispatchers
+
+### Perf
+
+- updated the memory injection warning so it only logs once, rather than after each keystroke
+
+## v0.7.4 (2026-07-02)
+
+### Feat
+
+- Pin specific usql version to sbx kit
+- recursively take ownership over the copied in coyote config for the sbx
+- explicitly specify the COYOTE_CONFIG_DIR in the sbx kit
+- --tail-logs can track log rollovers and incoporates a sleep timer to minimize idle CPU cycles
+- Added support for log rolling so log files don't just blow up over time
+
+### Fix
+
+- Added back in --kit specification for the running of the sbx
+- sbx isn't copying base files in their respective directories
+- Update deprecated sbx kit config
+- Properly chown the coyote config recursively and password file in the sbx
+
+## v0.7.3 (2026-06-24)
+
+### Fix
+
+- apply bootstrapping of functions at startup to fix edge case
+
+## v0.7.2 (2026-06-19)
+
+### Fix
+
+- usql version upgrade
+
+## v0.7.1 (2026-06-19)
+
+### Fix
+
+- sbx mixins must be passed in directories, not as files and the files must be named spec.yaml per new sbx version
+
+## v0.7.0 (2026-06-18)
+
+### Feat
+
+- added configurable cache path via the COYOTE_CACHE_PATH environment variable
+- added a memory option to .set tab completions
+- Added a diagnostic .info tools subcommand to make it easier to see what tools are enabled in all contexts
+- Added additional info outputs for enabled skills and sbx directories
+- directly execute shell commands from within the REPL
+- created mixin kit for built-in functions and MCP servers
+- Added sbx mixins for the secrets providers so users can also bootstrap those as well.
+- added support for loading sbx mixins that are dynamically discovered in the users workspace and config directory
+- Added a --fresh flag to let users create a truly bare bones sandbox without bootstrapping their config
+- initial built-in sandboxing support powered by Docker sbx
+- Added the ability to auto-bootstrap workspace memory when in git repos
+- Added explicit guardrail handling for pending agents
+- auto-append memory to memory index and don't necessarily require the LLM to remember to do it after a write
+- Added an --init-memory [global|workspace] flag to easily and quickly enable memory
+- added memory global configuration settings to the output of --info and .info
+- added .set memory REPL commands to control memory injection and applied formatting
+- Create the built-in memory management tools
+- Append the memory system prompts (readonly or r/w) to the system prompt when applicable
+- Created the --no-memory CLI flag to disable memory for this invocation
+- Added the memory configuration properties and storage to the main app config, roles, sessions, and agents.
+- initial scaffolding of a memory system
+
+### Fix
+
+- rebuild the tool scope after dynamically updating the skills_enabled value in the REPL
+- properly resolve Windows-based local vault password file locations and bootstrap them into the sandbox when possible
+- auto-translation of user-prefixed Mac and Linux paths for the vault password file when running inside a sandbox
+- don't attempt to auto complete .vault list in the REPL; that's the end of the command
+- buffer tool stdout as well as stderr so that any tools that error to stdout are captured and included in the response to the model, enabling the model to see what went wrong and to reason about how to fix it.
+- auto-bootstrapped memory was accidentally putting the MEMORY.md directly in the repo root rather than .coyote/memory/MEMORY.md
+- improved the fs_patch script description and added improved error handling to it.
+- added in forgotten require_max_tokens to the fable model
+- append memory functions to non-graph based agents on init
+- when auto_continue is disabled via the .set auto_continue false command, it should strip the todo functions from the list of functions
+- use rawPredict for non-streaming Claude requests
+
+### Refactor
+
+- Migrated the .skills command completion to use StateFlags and updated the help messages
+
+## v0.6.0 (2026-06-05)
+
+### Feat
+
+- added skill hint prompt injection and configuration
+- Fallthrough on missing secrets during mcp.json merging
+- validate visible_skills field at config load time
+- implemented reflexion (sorta) in sisyphus for significant code changes to delegate to the code-reviewer agent
+- improved explore agent
+- removed conditional fallback of LLM_*_RAW_JSON from built-ins
+- updated enabled_skills handling to support both list and comma-separated strings
+- added new REPL set commands for toggling skills and changing what skills are enabled
+- upgraded to the latest version of mcp-remote
+- fs_grep now works with both files and directories
+- improved code reviewer agents with skills
+- added round trip validation for vault providers to ensure permissions and authentication
+- created new first-time run wizard for secrets provider
+- vault_password_file or nothing at all is shorthand for just using the local gman provider for secret management
+- refactored gman usage to be generic and work with various vault providers and use the SupportedProvider enum directly for configurations
+- created initial parity gman generalization for vault provider
+- Refactored the sisyhpus agent system to utilize the new skills system to improve performance and reliability
+- llm graph nodes support skills
+- updated sisyphus and coder tools
+- removed potentially confusing tab completions for .skill
+- .edit skill <name> support from within the REPL
+- Added skills_dir to the info output of Coyote
+- Created a few auto built-in skills
+- Added support for auto_unload skills during chat
+- cleaned up skill implementation
+- support multiple skill flags to load multiple skills at CLI startup
+- Modified --skill CLI to allow users to specify skills to start the REPL or CLI with.
+- added CLI --skill flag for modifying skills easily
+- REPL integration with skills
+- dynamic loading/unloading of skill tools and MCP servers whenever load_skill/unload_skill are invoked
+- created built-in functions for listing, loading, and unloading skills
+- implemented the skills policy to track available skills per context
+- added remote install and install support for skills
+- created the skill registry
+- decided to make skills persist to disk like agents and not in-memory like built-in roles
+- scaffold skill module
+
+### Fix
+
+- disable skills for specific built-in roles
+- redirect stderr into user's /dev/tty for guards
+- azure doesn't support underscores in key vault
+- accidental regression on enabled_skills being empty = all
+- greedy secrets regex caused multiple secrets on one line to fail
+- add agent context check to skill visibility validation
+- enforced global visible_skills in llm node validation and improved skill loading error handling across the project
+- restore agent skill policy on error during effective policy calculation
+- apply the same validation for skill filenames on list_skills as happens everywhere else
+- the vault's init_bare should try to load the provisioned secret_provider from the config file without also interpolating any of the rest of the configuration file. It should only fail if the user has not yet created a configuration file; i.e. done a first-time run.
+- the vault roundtrip test used characters that are unsupported by some major secrets providers
+- fixed tool filtering logic for skills and user functions in agents
+- privilege leak when unloading skills and leaving tool scope untouched
+- When bootstrapping an app config to interpolate secrets, clone the secrets provider configuration as well so config secrets stored in remote vaults can be used properly
+- forgot to move back up the vault probe value error to be before the delete
+- don't silently fail on skill role composition extraction in llm nodes
+- set -euo pipefail for the temp script in execute_command.sh tool
+- added forgotten skill name validation to has_skill to prevent side-channel attacks
+- use unique values for the secrets round trip verification
+- stop interpolating a line if any errors occur
+- added path validation for skill names
+- effective_policy unconditionally overwrote skill values for role-like structs
+- updated execute_command to not mangle heredocs and also added explicit instructions to the coder and sisyphus agents to use fs_write and fs_patch over execute_command when writing files
+- llm nodes accidentally skipped skill_registry::effective_role because I was passing an inline role instead
+- updated temperature values for all agents and roles
+- added back in require_max_tokens for new Claude models
+- skill support also requires function calling to be enabled
+- non_tty tests break on some TTY terminals
+- skill loading on agents
+- forgot to bootstrap skills on REPL startup
+- remove now deprecated .skill edit command
+
+### Refactor
+
+- removed redundant skill name validation from has_skill function
+- support both CSV and list formats for enabled_tools
+- Support both CSV and list formats for enabled_mcp_servers
+
+## v0.5.0 (2026-05-27)
+
+### Feat
+
+- rename Loki to Coyote
+
+### Fix
+
+- bash-based user interactions in agents accidentally regressed in graph implementation
+- Claude function calling in agent contexts
+- Claude code rate limit error per new Claude changes
+
+## v0.4.0 (2026-05-23)
+
+### Feat
+
+- LLM node failures propgate up
+- Added .install remote tab completions to the REPL
+- feature complete install remote with category selection
+- Support to interactively add secrets to Coyote that are missing from MCP configs when merging
+- Added MCP config merging support for remote asset installations
+- install remote now writes files to disk
+- Created basic install_remote functions
+- Created a more comprehensive and immediately useful default config for first runs
+- Created an example graph-based agent called deep-research
+- Improved coder agent that is now a graph-based agent
+- Removed indicatif spinners. The UX just won't stop clobbering for parallel graph nodes
+- Added agent variables support for graph agents and improved script executor to use the same environment variables as normal agent tool calling for further flexibility
+- Improved UX with colored spinners for parallel graph agents and no clobbering outputs for sub-agents
+- created new graph-based deep-research agent
+- improved UX for parallel graph execution
+- added branch progress tracker for better visualization of parallel graph super-steps
+- Removed the jira-helper agent and replaced it with the atlassian role
+- created the RenderMode enum to suppress stdout streaming during parallel graph super-steps
+- Full support for map node types
+- implemented the frontier-based scheduling for the graph executor with simplified state management (gotta love .clone)
+- validation support for parallel graph execution; restricted map nodes to only run for nodes without next targets and not supporting chained map nodes
+- created the staging area for state merges per super-step and created the built-in reducers (and their application) for the state merge phase of a super step
+- scaffolding work for fan-out nodes for parallel branch execution support and stubbed out Map node types
+- Coyote can now update itself via .update and --update commands
+- added a .edit command for editing the MCP configuration file
+- Created a new .install command to install bundled assets on-demand
+- migrated llm node validation to graph loading time instead of graph runtime
+- ripped out user input timeout scaffolding for approval and input node types; implementation can't be done cleanly
+- added additional support for all RAG-configuration fields in RAG nodes
+- initial support for RAG nodes in the graph execution system
+- implemented structured logging for graph execution
+- merged normal agent config and graph agent configs into one file (either/or)
+- added structured-output extraction for llm and agent nodes
+- created full llm node runtime implementation
+- scaffolded together the initial llm node type and its executor
+- wired together graph execution and agent graph dispatch
+- implemented support for the graph executor
+- created the approval node executor and the input node executor for user interaction
+- Added initial support for native Coyote agent nodes in the graph-based agent system
+- Added direct script invocation support for graph-based agents
+- Added graph validation
+- Implemented state management for agent graphs
+- initial agent graph scaffolding
+- add auto-continue support to all contexts
+- dynamic tab completions now show the sessions for a given agent instead of only listing global sessions
+- legacy SSE support for MCP server configurations
+- support http/sse transport types for MCP server configurations so it fully supports claude desktop-style MCP configs
+- 99% complete migration to new state structs to get away from God-Config struct; i.e. AppConfig, AppState, and RequestContext
+- Automatic runtime customization using shebangs
+- Created a demo TypeScript tool and a get_current_weather function in TypeScript
+- Updated the Python demo tool to show all possible parameter types and variations
+- Added TypeScript tool support using the refactored common ScriptedLanguage trait
+
+### Fix
+
+- Generified the functions usage of script detection for an executable bit on unix systems
+- merge required claude code system prompt into instructions
+- updated argc argument passing in run-tool and run-agent scripts
+- Added additional graph validation for parallel reads and writes with dependencies between nodes states
+- bug in next_single method and improved outcome handling for LLM node execution
+- inline RAG bug when globbing files by extension without subdirectory globbing
+- update the estimate_token_length function to use the standard word count method
+- removed unnecessary regenerate logic for sessions and use the same logic for all contexts; prevents a panic on empty message list
+- error when users try to start a session on a graph agent
+- added on_other field for approval nodes so users can specify an alternative free-text target when none of the options match what they want
+- accidentally added back in full agent tools on LLM nodes
+- Improve the coder agent's usage of tools
+- make the agent__collect escalation-aware so it doesn't freeze on sub-agent escalations
+- check for an existing session before starting up MCP servers when switching to a role
+- do not switch to agent if a session is active.
+- Do not append todo instructions when function calling is disabled
+- a bug in the dynamic completions because the crate name is coyote-ai but the binary is named coyote
+- bug found by copilot that would create a lock on the PollSender for sse-based MCP servers
+- Accidental shadow of temp_file function for Windows function calling
+- upgraded to newer rmcp version to get native-tls support
+- RagCache was not being used for agent and sub-agent instantiation
+- TypeScript function args were being passed as objects rather than direct parameters
+- Added in forgotten wrapper scripts for TypeScript tools
+- don't shadow variables in binary path handling for Windows
+- Tool call improvements for Windows systems
+
+### Refactor
+
+- migrated llm nodes to use Roles to simplify instructions handling and to function like inline roles
+- migrated the next_node and apply_state_updates logic for LLM nodes into the LlmExecutor
+- fully complete state re-architecting
+- Fully ripped out the god Config struct
+- Deprecated old Config struct initialization logic
+- migrate functions and MCP servers to AppConfig
+- Migrate the vault/bare_init logic
+- created a single install_builtins free function to remove from Config::init
+- partial migration to init in AppConfig
+- Extracted common Python parser logic into a common.rs module
+- python tools now use tree-sitter queries instead of AST
+
 ## v0.3.0 (2026-04-02)
 
 ### Feat
@@ -21,7 +486,7 @@
 - Created a CodeRabbit-style code-reviewer agent
 - Added configuration option in agents to indicate the timeout for user input before proceeding (defaults to 5 minutes)
 - Added support for sub-agents to escalate user interaction requests from any depth to the parent agents for user interactions
-- built-in user interaction tools to remove the need for the list/confirm/etc prompts in prompt tools and to enhance user interactions in Loki
+- built-in user interaction tools to remove the need for the list/confirm/etc prompts in prompt tools and to enhance user interactions in Coyote
 - Experimental update to sisyphus to use the new parallel agent spawning system
 - Added an agent configuration property that allows auto-injecting sub-agent spawning instructions (when using the built-in sub-agent spawning system)
 - Auto-dispatch support of sub-agents and support for the teammate pattern between subagents
@@ -75,7 +540,7 @@
 
 - Simplified sisyphus prompt to improve functionality
 - Supported the injection of RAG sources into the prompt, not just via the `.sources rag` command in the REPL so models can directly reference the documents that supported their responses
-- Created the Sisyphus agent to make Loki function like Claude Code, Gemini, Codex, etc.
+- Created the Sisyphus agent to make Coyote function like Claude Code, Gemini, Codex, etc.
 - Created the Oracle agent to handle high-level architectural decisions and design questions about a given codebase
 - Updated the coder agent to be much more task-focused and to be delegated to by Sisyphus
 - Created the explore agent for exploring codebases to help answer questions
@@ -135,8 +600,8 @@
 - Support for secret injection into the global config file (API keys, for example)
 - Improved MCP handling toggle handling
 - Secret injection into the MCP configuration
-- added REPL support for interacting with the Loki vault
-- Integrated gman with Loki to create a vault and added flags to configure the Loki vault
+- added REPL support for interacting with the Coyote vault
+- Integrated gman with Coyote to create a vault and added flags to configure the Coyote vault
 - Added a default session to the jira helper to make interaction more natural
 - Created the repo-analyzer role
 - Created the coder and sql agents

CONTRIBUTING.md

@@ -2,7 +2,7 @@
 Contributors are very welcome! **No contribution is too small and all contributions are valued.**
 
 ## Rust
-You'll need to have the stable Rust toolchain installed in order to develop Loki.
+You'll need to have the stable Rust toolchain installed in order to develop Coyote.
 
 The Rust toolchain (stable) can be installed via rustup using the following command:
 
@@ -84,5 +84,5 @@ Claude, etc.) is not permitted unless explicitly disclosed and approved.
 Submissions must certify that the contributor understands and can maintain the code they submit.
 
 ## Questions? Reach out to me!
-If you encounter any questions while developing Loki, please don't hesitate to reach out to me at 
+If you encounter any questions while developing Coyote, please don't hesitate to reach out to me at 
 alex.j.tusa@gmail.com. I'm happy to help contributors in any way I can, regardless of if they're new or experienced!

CREDITS.md

@@ -1,19 +1,19 @@
 # Credits
 
 ## AIChat
-Loki originally started as a fork of the fantastic
+Coyote originally started as a fork of the fantastic
 [AIChat CLI](https://github.com/sigoden/aichat). The initial goal was simply
 to fix a bug in how MCP servers worked with AIChat, allowing different MCP
-servers to be specified per agent. Since then, Loki has evolved far beyond
+servers to be specified per agent. Since then, Coyote has evolved far beyond
 its original scope and grown into a passion project with a life of its own.
 
-Today, Loki includes first-class MCP server support (for both local and remote
+Today, Coyote includes first-class MCP server support (for both local and remote
 servers), a built-in vault for interpolating secrets in configuration files,
 built-in agents and macros, dynamic tab completions, integrated custom
 functions (no external `argc` dependency), improved documentation, and much
 more with many more ideas planned for the future.
 
-Loki is now developed and maintained as an independent project. Full credit
+Coyote is now developed and maintained as an independent project. Full credit
 for the original foundation goes to the developers of the wonderful
 AIChat project.
 
@@ -21,11 +21,20 @@ This project is not affiliated with or endorsed by the AIChat maintainers.
 
 ## AIChat
 
-Loki originally began as a fork of [AIChat CLI](https://github.com/sigoden/aichat),
+Coyote originally began as a fork of [AIChat CLI](https://github.com/sigoden/aichat),
 created and maintained by the AIChat contributors.
 
-While Loki has since diverged significantly and is now developed as an
+While Coyote has since diverged significantly and is now developed as an
 independent project, its early foundation and inspiration came from the
 AIChat project.
 
-AIChat is licensed under the MIT License.
+AIChat is licensed under the MIT License. The MIT license text and its
+copyright notice are preserved in the [LICENSE-MIT](./LICENSE-MIT) file.
+
+## Licensing
+
+Coyote as a whole is licensed under the GNU Affero General Public License
+v3.0 only (AGPL-3.0-only); see [LICENSE](./LICENSE). Substantial portions
+derived from AIChat remain under the MIT License (Copyright (c) sigoden),
+preserved in [LICENSE-MIT](./LICENSE-MIT). See [NOTICE](./NOTICE) for the
+combined-licensing summary.

Cargo.toml

@@ -1,23 +1,25 @@
 [package]
-name = "loki-ai"
-version = "0.3.0"
+name = "coyote-ai"
+version = "0.8.3"
 edition = "2024"
 authors = ["Alex Clarke <alex.j.tusa@gmail.com>"]
 description = "An all-in-one, batteries included LLM CLI Tool"
 keywords = ["chatgpt", "llm", "cli", "ai", "repl"]
-homepage = "https://github.com/Dark-Alex-17/loki"
-repository = "https://github.com/Dark-Alex-17/loki"
+homepage = "https://github.com/Dark-Alex-17/coyote"
+repository = "https://github.com/Dark-Alex-17/coyote"
 categories = ["command-line-utilities"]
 readme = "README.md"
-license = "MIT"
-rust-version = "1.89.0"
+license = "AGPL-3.0-only"
+rust-version = "1.95.0"
 exclude = [".github", "CONTRIBUTING.md"]
 
 [dependencies]
 anyhow = "1.0.69"
 bytes = "1.4.0"
 clap = { version = "4.5.40", features = ["cargo", "derive", "wrap_help"] }
+comfy-table = { version = "7.1.4", features = ["custom_styling"] }
 dirs = "6.0.0"
+duckdb = { version = "1.10505.0", features = ["bundled"] }
 dunce = "1.0.5"
 futures-util = "0.3.29"
 inquire = "0.9.4"
@@ -49,7 +51,7 @@ textwrap = "0.16.0"
 ansi_colours = "1.2.2"
 eventsource-stream = "0.2.3"
 log = "0.4.28"
-log4rs = { version = "1.4.0", features = ["file_appender"] }
+log4rs = { version = "1.4.0", features = ["file_appender", "rolling_file_appender", "compound_policy", "fixed_window_roller", "size_trigger"] }
 shell-words = "1.1.0"
 sha2 = "0.10.8"
 unicode-width = "0.2.0"
@@ -58,6 +60,8 @@ http = "1.1.0"
 indexmap = { version = "2.2.6", features = ["serde"] }
 hmac = "0.12.1"
 aws-smithy-eventstream = "0.60.4"
+aws-smithy-types = "=1.4.9"
+time = "=0.3.47"
 urlencoding = "2.1.3"
 json-patch = { version = "4.0.0", default-features = false }
 bitflags = "2.5.0"
@@ -72,6 +76,7 @@ html_to_markdown = "0.1.0"
 rust-embed = "8.5.0"
 os_info = { version = "3.8.2", default-features = false }
 bm25 = { version = "2.0.1", features = ["parallelism"] }
+petgraph = { version = "0.7", features = ["serde-1"] }
 which = "8.0.0"
 fuzzy-matcher = "0.3.7"
 terminal-colorsaurus = "0.4.8"
@@ -79,7 +84,7 @@ duct = "1.0.0"
 argc = "1.23.0"
 strum_macros = "0.27.2"
 indoc = "2.0.6"
-rmcp = { version = "1.5.0", features = [
+rmcp = { version = "3.1.2", features = [
   "client",
   "transport-child-process",
   "transport-streamable-http-client-reqwest",
@@ -91,7 +96,7 @@ tree-sitter-python = "0.25.0"
 tree-sitter-typescript = "0.23"
 colored = "3.0.0"
 clap_complete = { version = "4.5.58", features = ["unstable-dynamic"] }
-gman = "0.4.1"
+gman = "0.5.0"
 clap_complete_nushell = "4.5.9"
 open = "5"
 rand = { version = "0.10.0", features = ["default"] }
@@ -104,6 +109,7 @@ self_update = { version = "0.44", default-features = false, features = [
   "archive-zip",
   "compression-zip-deflate",
 ] }
+qrcode = "0.14"
 
 [dependencies.reqwest]
 version = "0.13.3"
@@ -138,10 +144,10 @@ pretty_assertions = "1.4.0"
 serial_test = "3"
 
 [[bin]]
-name = "loki"
+name = "coyote"
 path = "src/main.rs"
 
 [profile.release]
 lto = true
 strip = true
-opt-level = "z"
+opt-level = "z"

Dockerfile

@@ -0,0 +1,110 @@
+ARG COYOTE_VERSION
+FROM docker/sandbox-templates:shell-docker AS build
+
+ARG COYOTE_VERSION
+ARG TARGETARCH
+
+ENV PATH="/home/agent/.cargo/bin:/home/agent/.local/bin:${PATH}"
+
+USER root
+
+RUN apt-get update && \
+    apt-get install -y --no-install-recommends \
+      jq curl git \
+      build-essential pkg-config \
+      cmake \
+      clang libclang-dev \
+      musl-tools \
+      libssl-dev \
+      pandoc \
+      bzip2 \
+      nano && \
+    rm -rf /var/lib/apt/lists/*
+
+RUN set -euo pipefail; \
+    USQL_VERSION=0.21.4; \
+    case "${TARGETARCH}" in \
+      amd64) USQL_ARCH=amd64 ;; \
+      arm64) USQL_ARCH=arm64 ;; \
+      *) echo "Unsupported TARGETARCH: ${TARGETARCH}" >&2; exit 1 ;; \
+    esac; \
+    TMPDIR=$(mktemp -d); \
+    curl -fsSL --retry 3 \
+      "https://github.com/xo/usql/releases/download/v${USQL_VERSION}/usql_static-${USQL_VERSION}-linux-${USQL_ARCH}.tar.bz2" \
+      -o "$TMPDIR/usql.tar.bz2"; \
+    tar -xjf "$TMPDIR/usql.tar.bz2" -C "$TMPDIR"; \
+    install -m 0755 "$TMPDIR/usql_static" /usr/local/bin/usql; \
+    rm -rf "$TMPDIR"
+
+RUN set -euo pipefail; \
+    DUCKDB_VERSION=1.5.5; \
+    case "${TARGETARCH}" in \
+      amd64) DUCKDB_ARCH=amd64 ;; \
+      arm64) DUCKDB_ARCH=arm64 ;; \
+      *) echo "Unsupported TARGETARCH: ${TARGETARCH}" >&2; exit 1 ;; \
+    esac; \
+    TMPDIR=$(mktemp -d); \
+    curl -fsSL --retry 3 \
+      "https://github.com/duckdb/duckdb/releases/download/v${DUCKDB_VERSION}/duckdb_cli-linux-${DUCKDB_ARCH}.gz" \
+      -o "$TMPDIR/duckdb.gz"; \
+    gunzip "$TMPDIR/duckdb.gz"; \
+    install -m 0755 "$TMPDIR/duckdb" /usr/local/bin/duckdb; \
+    rm -rf "$TMPDIR"
+
+USER 1000
+
+RUN curl -LsSf https://astral.sh/uv/install.sh | sh && \
+    printf '#!/bin/sh\nexec uv tool run "$@"\n' > "$HOME/.local/bin/uvx" && \
+    chmod +x "$HOME/.local/bin/uvx"
+
+RUN mkdir -p /usr/local/share/npm-global/lib
+
+RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | \
+      sh -s -- -y --default-toolchain stable --profile minimal && \
+    . "$HOME/.cargo/env" && \
+    cargo install --locked iwec && \
+    cargo install --locked ast-grep
+
+USER root
+
+RUN set -euo pipefail; \
+    case "${TARGETARCH}" in \
+      amd64) MUSL_TARGET=x86_64-unknown-linux-musl ;; \
+      arm64) MUSL_TARGET=aarch64-unknown-linux-musl ;; \
+      *) echo "Unsupported TARGETARCH: ${TARGETARCH}" >&2; exit 1 ;; \
+    esac; \
+    TMPDIR=$(mktemp -d); \
+    curl -fsSL --retry 3 \
+      "https://github.com/Dark-Alex-17/coyote/releases/download/v${COYOTE_VERSION}/coyote-${MUSL_TARGET}.tar.gz" \
+      -o "$TMPDIR/coyote.tar.gz"; \
+    tar -xzf "$TMPDIR/coyote.tar.gz" -C "$TMPDIR"; \
+    install -m 0755 "$TMPDIR/coyote" /home/agent/.cargo/bin/coyote; \
+    chown 1000:1000 /home/agent/.cargo/bin/coyote; \
+    rm -rf "$TMPDIR"
+
+FROM scratch
+
+ARG COYOTE_VERSION
+
+COPY --from=build / /
+
+ENV PATH="/home/agent/.cargo/bin:/home/agent/.local/bin:/usr/local/share/npm-global/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin" \
+    NPM_CONFIG_PREFIX="/usr/local/share/npm-global" \
+    NO_PROXY="localhost,127.0.0.1,::1,172.17.0.0/16" \
+    no_proxy="localhost,127.0.0.1,::1,172.17.0.0/16" \
+    BASH_ENV="/etc/sandbox-persistent.sh"
+
+LABEL com.docker.sandboxes="templates" \
+      com.docker.sandboxes.base="ubuntu:questing" \
+      com.docker.sandboxes.flavor="shell-docker" \
+      com.docker.sandboxes.start-docker="true" \
+      org.opencontainers.image.title="coyote" \
+      org.opencontainers.image.description="An all-in-one, batteries-included LLM CLI tool: Shell Assistant, CLI & REPL mode, RAG, AI tools & agents, MCP servers, skills, and macros." \
+      org.opencontainers.image.source="https://github.com/Dark-Alex-17/coyote" \
+      org.opencontainers.image.version="${COYOTE_VERSION}"
+
+WORKDIR /home/agent/workspace
+
+USER 1000
+
+ENTRYPOINT ["coyote"]

LICENSE

@@ -1,22 +1,661 @@
-The MIT License (MIT)
+                    GNU AFFERO GENERAL PUBLIC LICENSE
+                       Version 3, 19 November 2007
 
-Copyright (c) 2025 sigoden
-Copyright (c) 2025 Alexander J. Clarke
+ Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
 
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
+                            Preamble
 
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
+  The GNU Affero General Public License is a free, copyleft license for
+software and other kinds of works, specifically designed to ensure
+cooperation with the community in the case of network server software.
 
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
+  The licenses for most software and other practical works are designed
+to take away your freedom to share and change the works.  By contrast,
+our General Public Licenses are intended to guarantee your freedom to
+share and change all versions of a program--to make sure it remains free
+software for all its users.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+them if you wish), that you receive source code or can get it if you
+want it, that you can change the software or use pieces of it in new
+free programs, and that you know you can do these things.
+
+  Developers that use our General Public Licenses protect your rights
+with two steps: (1) assert copyright on the software, and (2) offer
+you this License which gives you legal permission to copy, distribute
+and/or modify the software.
+
+  A secondary benefit of defending all users' freedom is that
+improvements made in alternate versions of the program, if they
+receive widespread use, become available for other developers to
+incorporate.  Many developers of free software are heartened and
+encouraged by the resulting cooperation.  However, in the case of
+software used on network servers, this result may fail to come about.
+The GNU General Public License permits making a modified version and
+letting the public access it on a server without ever releasing its
+source code to the public.
+
+  The GNU Affero General Public License is designed specifically to
+ensure that, in such cases, the modified source code becomes available
+to the community.  It requires the operator of a network server to
+provide the source code of the modified version running there to the
+users of that server.  Therefore, public use of a modified version, on
+a publicly accessible server, gives the public access to the source
+code of the modified version.
+
+  An older license, called the Affero General Public License and
+published by Affero, was designed to accomplish similar goals.  This is
+a different license, not a version of the Affero GPL, but Affero has
+released a new version of the Affero GPL which permits relicensing under
+this license.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+                       TERMS AND CONDITIONS
+
+  0. Definitions.
+
+  "This License" refers to version 3 of the GNU Affero General Public License.
+
+  "Copyright" also means copyright-like laws that apply to other kinds of
+works, such as semiconductor masks.
+
+  "The Program" refers to any copyrightable work licensed under this
+License.  Each licensee is addressed as "you".  "Licensees" and
+"recipients" may be individuals or organizations.
+
+  To "modify" a work means to copy from or adapt all or part of the work
+in a fashion requiring copyright permission, other than the making of an
+exact copy.  The resulting work is called a "modified version" of the
+earlier work or a work "based on" the earlier work.
+
+  A "covered work" means either the unmodified Program or a work based
+on the Program.
+
+  To "propagate" a work means to do anything with it that, without
+permission, would make you directly or secondarily liable for
+infringement under applicable copyright law, except executing it on a
+computer or modifying a private copy.  Propagation includes copying,
+distribution (with or without modification), making available to the
+public, and in some countries other activities as well.
+
+  To "convey" a work means any kind of propagation that enables other
+parties to make or receive copies.  Mere interaction with a user through
+a computer network, with no transfer of a copy, is not conveying.
+
+  An interactive user interface displays "Appropriate Legal Notices"
+to the extent that it includes a convenient and prominently visible
+feature that (1) displays an appropriate copyright notice, and (2)
+tells the user that there is no warranty for the work (except to the
+extent that warranties are provided), that licensees may convey the
+work under this License, and how to view a copy of this License.  If
+the interface presents a list of user commands or options, such as a
+menu, a prominent item in the list meets this criterion.
+
+  1. Source Code.
+
+  The "source code" for a work means the preferred form of the work
+for making modifications to it.  "Object code" means any non-source
+form of a work.
+
+  A "Standard Interface" means an interface that either is an official
+standard defined by a recognized standards body, or, in the case of
+interfaces specified for a particular programming language, one that
+is widely used among developers working in that language.
+
+  The "System Libraries" of an executable work include anything, other
+than the work as a whole, that (a) is included in the normal form of
+packaging a Major Component, but which is not part of that Major
+Component, and (b) serves only to enable use of the work with that
+Major Component, or to implement a Standard Interface for which an
+implementation is available to the public in source code form.  A
+"Major Component", in this context, means a major essential component
+(kernel, window system, and so on) of the specific operating system
+(if any) on which the executable work runs, or a compiler used to
+produce the work, or an object code interpreter used to run it.
+
+  The "Corresponding Source" for a work in object code form means all
+the source code needed to generate, install, and (for an executable
+work) run the object code and to modify the work, including scripts to
+control those activities.  However, it does not include the work's
+System Libraries, or general-purpose tools or generally available free
+programs which are used unmodified in performing those activities but
+which are not part of the work.  For example, Corresponding Source
+includes interface definition files associated with source files for
+the work, and the source code for shared libraries and dynamically
+linked subprograms that the work is specifically designed to require,
+such as by intimate data communication or control flow between those
+subprograms and other parts of the work.
+
+  The Corresponding Source need not include anything that users
+can regenerate automatically from other parts of the Corresponding
+Source.
+
+  The Corresponding Source for a work in source code form is that
+same work.
+
+  2. Basic Permissions.
+
+  All rights granted under this License are granted for the term of
+copyright on the Program, and are irrevocable provided the stated
+conditions are met.  This License explicitly affirms your unlimited
+permission to run the unmodified Program.  The output from running a
+covered work is covered by this License only if the output, given its
+content, constitutes a covered work.  This License acknowledges your
+rights of fair use or other equivalent, as provided by copyright law.
+
+  You may make, run and propagate covered works that you do not
+convey, without conditions so long as your license otherwise remains
+in force.  You may convey covered works to others for the sole purpose
+of having them make modifications exclusively for you, or provide you
+with facilities for running those works, provided that you comply with
+the terms of this License in conveying all material for which you do
+not control copyright.  Those thus making or running the covered works
+for you must do so exclusively on your behalf, under your direction
+and control, on terms that prohibit them from making any copies of
+your copyrighted material outside their relationship with you.
+
+  Conveying under any other circumstances is permitted solely under
+the conditions stated below.  Sublicensing is not allowed; section 10
+makes it unnecessary.
+
+  3. Protecting Users' Legal Rights From Anti-Circumvention Law.
+
+  No covered work shall be deemed part of an effective technological
+measure under any applicable law fulfilling obligations under article
+11 of the WIPO copyright treaty adopted on 20 December 1996, or
+similar laws prohibiting or restricting circumvention of such
+measures.
+
+  When you convey a covered work, you waive any legal power to forbid
+circumvention of technological measures to the extent such circumvention
+is effected by exercising rights under this License with respect to
+the covered work, and you disclaim any intention to limit operation or
+modification of the work as a means of enforcing, against the work's
+users, your or third parties' legal rights to forbid circumvention of
+technological measures.
+
+  4. Conveying Verbatim Copies.
+
+  You may convey verbatim copies of the Program's source code as you
+receive it, in any medium, provided that you conspicuously and
+appropriately publish on each copy an appropriate copyright notice;
+keep intact all notices stating that this License and any
+non-permissive terms added in accord with section 7 apply to the code;
+keep intact all notices of the absence of any warranty; and give all
+recipients a copy of this License along with the Program.
+
+  You may charge any price or no price for each copy that you convey,
+and you may offer support or warranty protection for a fee.
+
+  5. Conveying Modified Source Versions.
+
+  You may convey a work based on the Program, or the modifications to
+produce it from the Program, in the form of source code under the
+terms of section 4, provided that you also meet all of these conditions:
+
+    a) The work must carry prominent notices stating that you modified
+    it, and giving a relevant date.
+
+    b) The work must carry prominent notices stating that it is
+    released under this License and any conditions added under section
+    7.  This requirement modifies the requirement in section 4 to
+    "keep intact all notices".
+
+    c) You must license the entire work, as a whole, under this
+    License to anyone who comes into possession of a copy.  This
+    License will therefore apply, along with any applicable section 7
+    additional terms, to the whole of the work, and all its parts,
+    regardless of how they are packaged.  This License gives no
+    permission to license the work in any other way, but it does not
+    invalidate such permission if you have separately received it.
+
+    d) If the work has interactive user interfaces, each must display
+    Appropriate Legal Notices; however, if the Program has interactive
+    interfaces that do not display Appropriate Legal Notices, your
+    work need not make them do so.
+
+  A compilation of a covered work with other separate and independent
+works, which are not by their nature extensions of the covered work,
+and which are not combined with it such as to form a larger program,
+in or on a volume of a storage or distribution medium, is called an
+"aggregate" if the compilation and its resulting copyright are not
+used to limit the access or legal rights of the compilation's users
+beyond what the individual works permit.  Inclusion of a covered work
+in an aggregate does not cause this License to apply to the other
+parts of the aggregate.
+
+  6. Conveying Non-Source Forms.
+
+  You may convey a covered work in object code form under the terms
+of sections 4 and 5, provided that you also convey the
+machine-readable Corresponding Source under the terms of this License,
+in one of these ways:
+
+    a) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by the
+    Corresponding Source fixed on a durable physical medium
+    customarily used for software interchange.
+
+    b) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by a
+    written offer, valid for at least three years and valid for as
+    long as you offer spare parts or customer support for that product
+    model, to give anyone who possesses the object code either (1) a
+    copy of the Corresponding Source for all the software in the
+    product that is covered by this License, on a durable physical
+    medium customarily used for software interchange, for a price no
+    more than your reasonable cost of physically performing this
+    conveying of source, or (2) access to copy the
+    Corresponding Source from a network server at no charge.
+
+    c) Convey individual copies of the object code with a copy of the
+    written offer to provide the Corresponding Source.  This
+    alternative is allowed only occasionally and noncommercially, and
+    only if you received the object code with such an offer, in accord
+    with subsection 6b.
+
+    d) Convey the object code by offering access from a designated
+    place (gratis or for a charge), and offer equivalent access to the
+    Corresponding Source in the same way through the same place at no
+    further charge.  You need not require recipients to copy the
+    Corresponding Source along with the object code.  If the place to
+    copy the object code is a network server, the Corresponding Source
+    may be on a different server (operated by you or a third party)
+    that supports equivalent copying facilities, provided you maintain
+    clear directions next to the object code saying where to find the
+    Corresponding Source.  Regardless of what server hosts the
+    Corresponding Source, you remain obligated to ensure that it is
+    available for as long as needed to satisfy these requirements.
+
+    e) Convey the object code using peer-to-peer transmission, provided
+    you inform other peers where the object code and Corresponding
+    Source of the work are being offered to the general public at no
+    charge under subsection 6d.
+
+  A separable portion of the object code, whose source code is excluded
+from the Corresponding Source as a System Library, need not be
+included in conveying the object code work.
+
+  A "User Product" is either (1) a "consumer product", which means any
+tangible personal property which is normally used for personal, family,
+or household purposes, or (2) anything designed or sold for incorporation
+into a dwelling.  In determining whether a product is a consumer product,
+doubtful cases shall be resolved in favor of coverage.  For a particular
+product received by a particular user, "normally used" refers to a
+typical or common use of that class of product, regardless of the status
+of the particular user or of the way in which the particular user
+actually uses, or expects or is expected to use, the product.  A product
+is a consumer product regardless of whether the product has substantial
+commercial, industrial or non-consumer uses, unless such uses represent
+the only significant mode of use of the product.
+
+  "Installation Information" for a User Product means any methods,
+procedures, authorization keys, or other information required to install
+and execute modified versions of a covered work in that User Product from
+a modified version of its Corresponding Source.  The information must
+suffice to ensure that the continued functioning of the modified object
+code is in no case prevented or interfered with solely because
+modification has been made.
+
+  If you convey an object code work under this section in, or with, or
+specifically for use in, a User Product, and the conveying occurs as
+part of a transaction in which the right of possession and use of the
+User Product is transferred to the recipient in perpetuity or for a
+fixed term (regardless of how the transaction is characterized), the
+Corresponding Source conveyed under this section must be accompanied
+by the Installation Information.  But this requirement does not apply
+if neither you nor any third party retains the ability to install
+modified object code on the User Product (for example, the work has
+been installed in ROM).
+
+  The requirement to provide Installation Information does not include a
+requirement to continue to provide support service, warranty, or updates
+for a work that has been modified or installed by the recipient, or for
+the User Product in which it has been modified or installed.  Access to a
+network may be denied when the modification itself materially and
+adversely affects the operation of the network or violates the rules and
+protocols for communication across the network.
+
+  Corresponding Source conveyed, and Installation Information provided,
+in accord with this section must be in a format that is publicly
+documented (and with an implementation available to the public in
+source code form), and must require no special password or key for
+unpacking, reading or copying.
+
+  7. Additional Terms.
+
+  "Additional permissions" are terms that supplement the terms of this
+License by making exceptions from one or more of its conditions.
+Additional permissions that are applicable to the entire Program shall
+be treated as though they were included in this License, to the extent
+that they are valid under applicable law.  If additional permissions
+apply only to part of the Program, that part may be used separately
+under those permissions, but the entire Program remains governed by
+this License without regard to the additional permissions.
+
+  When you convey a copy of a covered work, you may at your option
+remove any additional permissions from that copy, or from any part of
+it.  (Additional permissions may be written to require their own
+removal in certain cases when you modify the work.)  You may place
+additional permissions on material, added by you to a covered work,
+for which you have or can give appropriate copyright permission.
+
+  Notwithstanding any other provision of this License, for material you
+add to a covered work, you may (if authorized by the copyright holders of
+that material) supplement the terms of this License with terms:
+
+    a) Disclaiming warranty or limiting liability differently from the
+    terms of sections 15 and 16 of this License; or
+
+    b) Requiring preservation of specified reasonable legal notices or
+    author attributions in that material or in the Appropriate Legal
+    Notices displayed by works containing it; or
+
+    c) Prohibiting misrepresentation of the origin of that material, or
+    requiring that modified versions of such material be marked in
+    reasonable ways as different from the original version; or
+
+    d) Limiting the use for publicity purposes of names of licensors or
+    authors of the material; or
+
+    e) Declining to grant rights under trademark law for use of some
+    trade names, trademarks, or service marks; or
+
+    f) Requiring indemnification of licensors and authors of that
+    material by anyone who conveys the material (or modified versions of
+    it) with contractual assumptions of liability to the recipient, for
+    any liability that these contractual assumptions directly impose on
+    those licensors and authors.
+
+  All other non-permissive additional terms are considered "further
+restrictions" within the meaning of section 10.  If the Program as you
+received it, or any part of it, contains a notice stating that it is
+governed by this License along with a term that is a further
+restriction, you may remove that term.  If a license document contains
+a further restriction but permits relicensing or conveying under this
+License, you may add to a covered work material governed by the terms
+of that license document, provided that the further restriction does
+not survive such relicensing or conveying.
+
+  If you add terms to a covered work in accord with this section, you
+must place, in the relevant source files, a statement of the
+additional terms that apply to those files, or a notice indicating
+where to find the applicable terms.
+
+  Additional terms, permissive or non-permissive, may be stated in the
+form of a separately written license, or stated as exceptions;
+the above requirements apply either way.
+
+  8. Termination.
+
+  You may not propagate or modify a covered work except as expressly
+provided under this License.  Any attempt otherwise to propagate or
+modify it is void, and will automatically terminate your rights under
+this License (including any patent licenses granted under the third
+paragraph of section 11).
+
+  However, if you cease all violation of this License, then your
+license from a particular copyright holder is reinstated (a)
+provisionally, unless and until the copyright holder explicitly and
+finally terminates your license, and (b) permanently, if the copyright
+holder fails to notify you of the violation by some reasonable means
+prior to 60 days after the cessation.
+
+  Moreover, your license from a particular copyright holder is
+reinstated permanently if the copyright holder notifies you of the
+violation by some reasonable means, this is the first time you have
+received notice of violation of this License (for any work) from that
+copyright holder, and you cure the violation prior to 30 days after
+your receipt of the notice.
+
+  Termination of your rights under this section does not terminate the
+licenses of parties who have received copies or rights from you under
+this License.  If your rights have been terminated and not permanently
+reinstated, you do not qualify to receive new licenses for the same
+material under section 10.
+
+  9. Acceptance Not Required for Having Copies.
+
+  You are not required to accept this License in order to receive or
+run a copy of the Program.  Ancillary propagation of a covered work
+occurring solely as a consequence of using peer-to-peer transmission
+to receive a copy likewise does not require acceptance.  However,
+nothing other than this License grants you permission to propagate or
+modify any covered work.  These actions infringe copyright if you do
+not accept this License.  Therefore, by modifying or propagating a
+covered work, you indicate your acceptance of this License to do so.
+
+  10. Automatic Licensing of Downstream Recipients.
+
+  Each time you convey a covered work, the recipient automatically
+receives a license from the original licensors, to run, modify and
+propagate that work, subject to this License.  You are not responsible
+for enforcing compliance by third parties with this License.
+
+  An "entity transaction" is a transaction transferring control of an
+organization, or substantially all assets of one, or subdividing an
+organization, or merging organizations.  If propagation of a covered
+work results from an entity transaction, each party to that
+transaction who receives a copy of the work also receives whatever
+licenses to the work the party's predecessor in interest had or could
+give under the previous paragraph, plus a right to possession of the
+Corresponding Source of the work from the predecessor in interest, if
+the predecessor has it or can get it with reasonable efforts.
+
+  You may not impose any further restrictions on the exercise of the
+rights granted or affirmed under this License.  For example, you may
+not impose a license fee, royalty, or other charge for exercise of
+rights granted under this License, and you may not initiate litigation
+(including a cross-claim or counterclaim in a lawsuit) alleging that
+any patent claim is infringed by making, using, selling, offering for
+sale, or importing the Program or any portion of it.
+
+  11. Patents.
+
+  A "contributor" is a copyright holder who authorizes use under this
+License of the Program or a work on which the Program is based.  The
+work thus licensed is called the contributor's "contributor version".
+
+  A contributor's "essential patent claims" are all patent claims
+owned or controlled by the contributor, whether already acquired or
+hereafter acquired, that would be infringed by some manner, permitted
+by this License, of making, using, or selling its contributor version,
+but do not include claims that would be infringed only as a
+consequence of further modification of the contributor version.  For
+purposes of this definition, "control" includes the right to grant
+patent sublicenses in a manner consistent with the requirements of
+this License.
+
+  Each contributor grants you a non-exclusive, worldwide, royalty-free
+patent license under the contributor's essential patent claims, to
+make, use, sell, offer for sale, import and otherwise run, modify and
+propagate the contents of its contributor version.
+
+  In the following three paragraphs, a "patent license" is any express
+agreement or commitment, however denominated, not to enforce a patent
+(such as an express permission to practice a patent or covenant not to
+sue for patent infringement).  To "grant" such a patent license to a
+party means to make such an agreement or commitment not to enforce a
+patent against the party.
+
+  If you convey a covered work, knowingly relying on a patent license,
+and the Corresponding Source of the work is not available for anyone
+to copy, free of charge and under the terms of this License, through a
+publicly available network server or other readily accessible means,
+then you must either (1) cause the Corresponding Source to be so
+available, or (2) arrange to deprive yourself of the benefit of the
+patent license for this particular work, or (3) arrange, in a manner
+consistent with the requirements of this License, to extend the patent
+license to downstream recipients.  "Knowingly relying" means you have
+actual knowledge that, but for the patent license, your conveying the
+covered work in a country, or your recipient's use of the covered work
+in a country, would infringe one or more identifiable patents in that
+country that you have reason to believe are valid.
+
+  If, pursuant to or in connection with a single transaction or
+arrangement, you convey, or propagate by procuring conveyance of, a
+covered work, and grant a patent license to some of the parties
+receiving the covered work authorizing them to use, propagate, modify
+or convey a specific copy of the covered work, then the patent license
+you grant is automatically extended to all recipients of the covered
+work and works based on it.
+
+  A patent license is "discriminatory" if it does not include within
+the scope of its coverage, prohibits the exercise of, or is
+conditioned on the non-exercise of one or more of the rights that are
+specifically granted under this License.  You may not convey a covered
+work if you are a party to an arrangement with a third party that is
+in the business of distributing software, under which you make payment
+to the third party based on the extent of your activity of conveying
+the work, and under which the third party grants, to any of the
+parties who would receive the covered work from you, a discriminatory
+patent license (a) in connection with copies of the covered work
+conveyed by you (or copies made from those copies), or (b) primarily
+for and in connection with specific products or compilations that
+contain the covered work, unless you entered into that arrangement,
+or that patent license was granted, prior to 28 March 2007.
+
+  Nothing in this License shall be construed as excluding or limiting
+any implied license or other defenses to infringement that may
+otherwise be available to you under applicable patent law.
+
+  12. No Surrender of Others' Freedom.
+
+  If conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot convey a
+covered work so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you may
+not convey it at all.  For example, if you agree to terms that obligate you
+to collect a royalty for further conveying from those to whom you convey
+the Program, the only way you could satisfy both those terms and this
+License would be to refrain entirely from conveying the Program.
+
+  13. Remote Network Interaction; Use with the GNU General Public License.
+
+  Notwithstanding any other provision of this License, if you modify the
+Program, your modified version must prominently offer all users
+interacting with it remotely through a computer network (if your version
+supports such interaction) an opportunity to receive the Corresponding
+Source of your version by providing access to the Corresponding Source
+from a network server at no charge, through some standard or customary
+means of facilitating copying of software.  This Corresponding Source
+shall include the Corresponding Source for any work covered by version 3
+of the GNU General Public License that is incorporated pursuant to the
+following paragraph.
+
+  Notwithstanding any other provision of this License, you have
+permission to link or combine any covered work with a work licensed
+under version 3 of the GNU General Public License into a single
+combined work, and to convey the resulting work.  The terms of this
+License will continue to apply to the part which is the covered work,
+but the work with which it is combined will remain governed by version
+3 of the GNU General Public License.
+
+  14. Revised Versions of this License.
+
+  The Free Software Foundation may publish revised and/or new versions of
+the GNU Affero General Public License from time to time.  Such new versions
+will be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+  Each version is given a distinguishing version number.  If the
+Program specifies that a certain numbered version of the GNU Affero General
+Public License "or any later version" applies to it, you have the
+option of following the terms and conditions either of that numbered
+version or of any later version published by the Free Software
+Foundation.  If the Program does not specify a version number of the
+GNU Affero General Public License, you may choose any version ever published
+by the Free Software Foundation.
+
+  If the Program specifies that a proxy can decide which future
+versions of the GNU Affero General Public License can be used, that proxy's
+public statement of acceptance of a version permanently authorizes you
+to choose that version for the Program.
+
+  Later license versions may give you additional or different
+permissions.  However, no additional obligations are imposed on any
+author or copyright holder as a result of your choosing to follow a
+later version.
+
+  15. Disclaimer of Warranty.
+
+  THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
+APPLICABLE LAW.  EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
+HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
+OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
+THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
+IS WITH YOU.  SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
+ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+  16. Limitation of Liability.
+
+  IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
+THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
+GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
+USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
+DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
+PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
+EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
+SUCH DAMAGES.
+
+  17. Interpretation of Sections 15 and 16.
+
+  If the disclaimer of warranty and limitation of liability provided
+above cannot be given local legal effect according to their terms,
+reviewing courts shall apply local law that most closely approximates
+an absolute waiver of all civil liability in connection with the
+Program, unless a warranty or assumption of liability accompanies a
+copy of the Program in return for a fee.
+
+                     END OF TERMS AND CONDITIONS
+
+            How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+state the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the program's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU Affero General Public License as published by
+    the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU Affero General Public License for more details.
+
+    You should have received a copy of the GNU Affero General Public License
+    along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+Also add information on how to contact you by electronic and paper mail.
+
+  If your software can interact with users remotely through a computer
+network, you should also make sure that it provides a way for users to
+get its source.  For example, if your program is a web application, its
+interface could display a "Source" link that leads users to an archive
+of the code.  There are many ways you could offer source, and different
+solutions will be better for different programs; see section 13 for the
+specific requirements.
+
+  You should also get your employer (if you work as a programmer) or school,
+if any, to sign a "copyright disclaimer" for the program, if necessary.
+For more information on this, and how to apply and follow the GNU AGPL, see
+<https://www.gnu.org/licenses/>.

LICENSE-MIT

@@ -0,0 +1,22 @@
+The MIT License (MIT)
+
+Copyright (c) 2025 sigoden
+Copyright (c) 2025 Alexander J. Clarke
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

NOTICE

@@ -0,0 +1,24 @@
+Coyote
+Copyright (c) 2025 Alexander J. Clarke
+
+This project as a whole is licensed under the GNU Affero General Public
+License, version 3.0 only (AGPL-3.0-only). The full text of that license is
+provided in the LICENSE file.
+
+--------------------------------------------------------------------------------
+Upstream / third-party notices
+--------------------------------------------------------------------------------
+
+Coyote began as a fork of AIChat (https://github.com/sigoden/aichat),
+Copyright (c) sigoden, which is distributed under the MIT License. Substantial
+portions of Coyote are derived from AIChat and remain available under the terms
+of the MIT License. The MIT License text and its required copyright and
+permission notices are preserved in the LICENSE-MIT file.
+
+As permitted by the MIT License, these portions have been incorporated into a
+larger work that is distributed under the AGPL-3.0-only license. When you
+receive Coyote as a combined work, your rights and obligations for the work as
+a whole are governed by the AGPL-3.0-only license; the MIT notice is retained
+to satisfy the attribution requirements of the MIT-licensed portions.
+
+See CREDITS.md for additional background and attribution.

README.md

@@ -1,54 +1,60 @@
-# Loki: All-in-one, batteries-included LLM CLI Tool
+# Coyote: All-in-one, batteries-included LLM CLI Tool
 
-![Test](https://github.com/Dark-Alex-17/loki/actions/workflows/ci.yaml/badge.svg)
-[![crates.io link](https://img.shields.io/crates/v/loki-ai.svg)](https://crates.io/crates/loki-ai)
-![Release](https://img.shields.io/github/v/release/Dark-Alex-17/loki?color=%23c694ff)
-![Crate.io downloads](https://img.shields.io/crates/d/loki-ai?label=Crate%20downloads)
-[![GitHub Downloads](https://img.shields.io/github/downloads/Dark-Alex-17/loki/total.svg?label=GitHub%20downloads)](https://github.com/Dark-Alex-17/loki/releases)
+![Test](https://github.com/Dark-Alex-17/coyote/actions/workflows/ci.yaml/badge.svg)
+[![crates.io link](https://img.shields.io/crates/v/coyote-ai.svg)](https://crates.io/crates/coyote-ai)
+![Release](https://img.shields.io/github/v/release/Dark-Alex-17/coyote?color=%23c694ff)
+![Crate.io downloads](https://img.shields.io/crates/d/coyote-ai?label=Crate%20downloads)
+[![GitHub Downloads](https://img.shields.io/github/downloads/Dark-Alex-17/coyote/total.svg?label=GitHub%20downloads)](https://github.com/Dark-Alex-17/coyote/releases)
+![Docker pulls](https://img.shields.io/docker/pulls/darkalex17/coyote?label=Docker%20downloads)
+[![License: AGPL v3](https://img.shields.io/badge/License-AGPL_v3-blue.svg)](./LICENSE)
 
-Loki is an all-in-one, batteries-included, LLM CLI tool featuring Shell Assistant, CLI & REPL Mode, RAG, AI Tools & 
+Coyote is an all-in-one, batteries-included, LLM CLI tool featuring Shell Assistant, CLI & REPL Mode, RAG, AI Tools & 
 Agents, and More.
 
-It is designed to include a number of useful agents, roles, macros, and more so users can get up and running with Loki 
+It is designed to include a number of useful agents, roles, macros, and more so users can get up and running with Coyote 
 in as little time as possible. You can also install entire bundles of agents, roles, macros, tools, and MCP servers from 
-any git repository — see [Sharing Configurations](#sharing-configurations).
+any git repository. See [Sharing Configurations](https://github.com/Dark-Alex-17/coyote/wiki/Sharing-Configurations) for more information.
 
-![Agent example](https://raw.githubusercontent.com/wiki/Dark-Alex-17/loki/images/agents/sql.gif)
+![Agent example](https://raw.githubusercontent.com/wiki/Dark-Alex-17/coyote/images/agents/sql.gif)
 
-Coming from [AIChat](https://github.com/sigoden/aichat)? Follow the [migration guide](https://github.com/Dark-Alex-17/loki/wiki/AIChat-Migration) to get started.
+Coming from [AIChat](https://github.com/sigoden/aichat)? Follow the [migration guide](https://github.com/Dark-Alex-17/coyote/wiki/AIChat-Migration) to get started.
 
 ## Quick Links
-* [AIChat Migration Guide](https://github.com/Dark-Alex-17/loki/wiki/AIChat-Migration): Coming from AIChat? Follow the migration guide to get started.
-* [Installation](#install): Install Loki
-* [Getting Started](#getting-started): Get started with Loki by doing first-run setup steps.
-* [Sharing Configurations](https://github.com/Dark-Alex-17/loki/wiki/Sharing-Configurations): Install bundles of agents, roles, macros, tools, and MCP servers from any git repo, and share your own.
-* [REPL](https://github.com/Dark-Alex-17/loki/wiki/REPL): Interactive Read-Eval-Print Loop for conversational interactions with LLMs and Loki.
-  * [Custom REPL Prompt](https://github.com/Dark-Alex-17/loki/wiki/REPL-Prompt): Customize the REPL prompt to provide useful contextual information.
-* [Vault](https://github.com/Dark-Alex-17/loki/wiki/Vault): Securely store and manage sensitive information such as API keys and credentials.
-* [Shell Integrations](https://github.com/Dark-Alex-17/loki/wiki/Shell-Integrations): Seamlessly integrate Loki with your shell environment for enhanced command-line assistance.
-* [Function Calling](https://github.com/Dark-Alex-17/loki/wiki/Tools): Leverage function calling capabilities to extend Loki's functionality with custom tools
-    * [Creating Custom Tools](https://github.com/Dark-Alex-17/loki/wiki/Custom-Tools): You can create your own custom tools to enhance Loki's capabilities.
-        * [Create Custom Python Tools](https://github.com/Dark-Alex-17/loki/wiki/Custom-Tools#custom-python-based-tools)
-        * [Create Custom TypeScript Tools](https://github.com/Dark-Alex-17/loki/wiki/Custom-Tools#custom-typescript-based-tools)
-        * [Create Custom Bash Tools](https://github.com/Dark-Alex-17/loki/wiki/Custom-Bash-Tools)
-            * [Bash Prompt Utilities](https://github.com/Dark-Alex-17/loki/wiki/Bash-Prompt-Helpers)
-* [First-Class MCP Server Support](https://github.com/Dark-Alex-17/loki/wiki/MCP-Servers): Easily connect and interact with MCP servers for advanced functionality.
-* [Macros](https://github.com/Dark-Alex-17/loki/wiki/Macros): Automate repetitive tasks and workflows with Loki "scripts" (macros).
-* [RAG](https://github.com/Dark-Alex-17/loki/wiki/RAG): Retrieval-Augmented Generation for enhanced information retrieval and generation.
-* [Sessions](https://github.com/Dark-Alex-17/loki/wiki/Sessions): Manage and persist conversational contexts and settings across multiple interactions.
-* [Roles](https://github.com/Dark-Alex-17/loki/wiki/Roles): Customize model behavior for specific tasks or domains.
-* [Agents](https://github.com/Dark-Alex-17/loki/wiki/Agents): Leverage AI agents to perform complex tasks and workflows, including sub-agent spawning, teammate messaging, and user interaction tools.
-    * [Graph Agents](https://github.com/Dark-Alex-17/loki/wiki/Graph-Agents): Define an agent as a declarative, YAML-driven workflow. A directed graph of typed nodes (LLM calls, scripts, approvals, user input, RAG retrieval, sub-agent spawns).
-* [Todo System](https://github.com/Dark-Alex-17/loki/wiki/TODO-System): Built-in task tracking for improved LLM reliability with smaller models.
-* [Environment Variables](https://github.com/Dark-Alex-17/loki/wiki/Environment-Variables): Override and customize your Loki configuration at runtime with environment variables.
-* [Client Configurations](https://github.com/Dark-Alex-17/loki/wiki/Clients): Configuration instructions for various LLM providers.
-    * [Authentication (API Key & OAuth)](https://github.com/Dark-Alex-17/loki/wiki/Clients#authentication): Authenticate with API keys or OAuth for subscription-based access.
-    * [Patching API Requests](https://github.com/Dark-Alex-17/loki/wiki/Patches): Learn how to patch API requests for advanced customization.
-* [Custom Themes](https://github.com/Dark-Alex-17/loki/wiki/Themes): Change the look and feel of Loki to your preferences with custom themes.
-* [History](#history): A history of how Loki came to be.
+* [AIChat Migration Guide](https://github.com/Dark-Alex-17/coyote/wiki/AIChat-Migration): Coming from AIChat? Follow the migration guide to get started.
+* [Installation](#install): Install Coyote
+* [Getting Started](#getting-started): Get started with Coyote by doing first-run setup steps.
+* [Sharing Configurations](https://github.com/Dark-Alex-17/coyote/wiki/Sharing-Configurations): Install bundles of agents, roles, macros, tools, and MCP servers from any git repo, and share your own.
+* [REPL](https://github.com/Dark-Alex-17/coyote/wiki/REPL): Interactive Read-Eval-Print Loop for conversational interactions with LLMs and Coyote.
+  * [Custom REPL Prompt](https://github.com/Dark-Alex-17/coyote/wiki/REPL-Prompt): Customize the REPL prompt to provide useful contextual information.
+* [Vault](https://github.com/Dark-Alex-17/coyote/wiki/Vault): Securely store and manage sensitive information such as API keys and credentials.
+* [Sandboxes](https://github.com/Dark-Alex-17/coyote/wiki/Sandboxes): Launch Coyote inside an isolated [Docker Sandbox](https://docs.docker.com/ai/sandboxes/) with one command. Host config and vault credentials are projected in automatically; everything else is delegated to the `sbx` CLI.
+* [Shell Integrations](https://github.com/Dark-Alex-17/coyote/wiki/Shell-Integrations): Seamlessly integrate Coyote with your shell environment for enhanced command-line assistance.
+* [Function Calling](https://github.com/Dark-Alex-17/coyote/wiki/Tools): Leverage function calling capabilities to extend Coyote's functionality with custom tools
+    * [Creating Custom Tools](https://github.com/Dark-Alex-17/coyote/wiki/Custom-Tools): You can create your own custom tools to enhance Coyote's capabilities.
+        * [Create Custom Python Tools](https://github.com/Dark-Alex-17/coyote/wiki/Custom-Tools#custom-python-based-tools)
+        * [Create Custom TypeScript Tools](https://github.com/Dark-Alex-17/coyote/wiki/Custom-Tools#custom-typescript-based-tools)
+        * [Create Custom Bash Tools](https://github.com/Dark-Alex-17/coyote/wiki/Custom-Bash-Tools)
+            * [Bash Prompt Utilities](https://github.com/Dark-Alex-17/coyote/wiki/Bash-Prompt-Helpers)
+* [First-Class MCP Server Support](https://github.com/Dark-Alex-17/coyote/wiki/MCP-Servers): Easily connect and interact with MCP servers for advanced functionality.
+* [Macros](https://github.com/Dark-Alex-17/coyote/wiki/Macros): Automate repetitive tasks and workflows with Coyote "scripts" (macros).
+* [RAG](https://github.com/Dark-Alex-17/coyote/wiki/RAG): Retrieval-Augmented Generation for enhanced information retrieval and generation.
+* [Sessions](https://github.com/Dark-Alex-17/coyote/wiki/Sessions): Manage and persist conversational contexts and settings across multiple interactions.
+* [Memory](https://github.com/Dark-Alex-17/coyote/wiki/Memory): Persistent file-based memory that survives across sessions. Bootstrap with `coyote --init-memory [global|workspace]`.
+* [Workspace Instructions](https://github.com/Dark-Alex-17/coyote/wiki/Workspace-Instructions): Human-curated project instructions (`COYOTE.md`) injected into every prompt, with `AGENTS.md`/`CLAUDE.md`/`GEMINI.md` fallbacks for cross-tool compatibility. Scaffold with `coyote --init-instructions`.
+* [Roles](https://github.com/Dark-Alex-17/coyote/wiki/Roles): Customize model behavior for specific tasks or domains.
+* [Skills](https://github.com/Dark-Alex-17/coyote/wiki/Skills): Modular knowledge or capability packs the LLM can load and unload mid-conversation. Multiple skills compose; instructions stack, tools and MCPs union.
+* [Agents](https://github.com/Dark-Alex-17/coyote/wiki/Agents): Leverage AI agents to perform complex tasks and workflows, including sub-agent spawning, teammate messaging, and user interaction tools.
+    * [Graph Agents](https://github.com/Dark-Alex-17/coyote/wiki/Graph-Agents): Define an agent as a declarative, YAML-driven workflow. A directed graph of typed nodes (LLM calls, scripts, approvals, user input, RAG retrieval, sub-agent spawns).
+* [Todo System](https://github.com/Dark-Alex-17/coyote/wiki/TODO-System): Built-in task tracking for improved LLM reliability with smaller models.
+* [Environment Variables](https://github.com/Dark-Alex-17/coyote/wiki/Environment-Variables): Override and customize your Coyote configuration at runtime with environment variables.
+* [Client Configurations](https://github.com/Dark-Alex-17/coyote/wiki/Clients): Configuration instructions for various LLM providers.
+    * [Authentication (API Key & OAuth)](https://github.com/Dark-Alex-17/coyote/wiki/Clients#authentication): Authenticate with API keys or OAuth for subscription-based access.
+    * [Patching API Requests](https://github.com/Dark-Alex-17/coyote/wiki/Patches): Learn how to patch API requests for advanced customization.
+* [Custom Themes](https://github.com/Dark-Alex-17/coyote/wiki/Themes): Change the look and feel of Coyote to your preferences with custom themes.
+* [History](#history): A history of how Coyote came to be.
 
 ## Prerequisites
-Loki requires the following tools to be installed on your system:
+Coyote requires the following tools to be installed on your system:
 * [jq](https://github.com/jqlang/jq)
     * `brew install jq`
 * [usql](https://github.com/xo/usql) (For the `sql` agent)
@@ -56,58 +62,94 @@ Loki requires the following tools to be installed on your system:
 * [docker](https://docs.docker.com/engine/install/)
 * [uv](https://docs.astral.sh/uv/getting-started/installation/)
     * `curl -LsSf https://astral.sh/uv/install.sh | sh`
+* [iwe](https://github.com/iwe-org/iwe) (`iwec`, for the built-in `iwe` MCP server that navigates large markdown knowledgebases)
+    * **Homebrew:** `brew tap iwe-org/iwe && brew trust --formula iwe-org/iwe/iwe && brew install iwe`
+    * **Cargo:** `cargo install iwec`
+* [ast-grep](https://ast-grep.github.io/) (for the built-in `ast_grep` structural code search tool, used by the `explore` agent)
+    * **Homebrew:** `brew install ast-grep`
+    * **Cargo:** `cargo install ast-grep --locked`
+    * **npm:** `npm i -g @ast-grep/cli`
+    * Optional: if `ast-grep` is not installed, the `ast_grep` tool reports it and agents fall back to `fs_grep`
+* [duckdb](https://duckdb.org/) (for fast, local RAGs)
+    * `curl https://install.duckdb.org | sh`
 
-These tools are used to provide various functionalities within Loki, such as document processing, JSON manipulation,
+These tools are used to provide various functionalities within Coyote, such as document processing, JSON manipulation,
 etc., and they are used within agents and tools.
 
 ## Install
 
 ### Cargo
-If you have Cargo installed, then you can install `loki` from Crates.io:
+If you have Cargo installed, then you can install `coyote` from Crates.io:
 
 ```shell
-cargo install loki-ai # Binary name is `loki`
+cargo install coyote-ai # Binary name is `coyote`
 
 # If you encounter issues installing, try installing with '--locked'
-cargo install --locked loki-ai
+cargo install --locked coyote-ai
 ```
 
 ### Homebrew (Mac/Linux)
-To install Loki from Homebrew, install the `loki` tap. Then you'll be able to install `loki`:
+To install Coyote from Homebrew, install the `coyote` tap. Then you'll be able to install `coyote`:
 
 ```shell
-brew tap Dark-Alex-17/loki
-brew install loki
+brew tap Dark-Alex-17/coyote
+brew install coyote
 
 # If you need to be more specific, use:
-brew install Dark-Alex-17/loki/loki
+brew install Dark-Alex-17/coyote/coyote
 ```
 
-To upgrade `loki` using Homebrew:
+To upgrade `coyote` using Homebrew:
 
 ```shell
-brew upgrade loki
+brew upgrade coyote
+```
+
+### Docker
+Coyote is available as a Docker image on Docker Hub (`darkalex17/coyote`) for Linux amd64 and arm64.
+Useful for CI, ephemeral environments, or anywhere you prefer not to install it natively.
+
+```bash
+docker pull darkalex17/coyote
+docker run --rm -it darkalex17/coyote
+```
+
+To persist your configuration across container runs, mount your existing config directory:
+
+```bash
+docker run --rm -it \
+  -v ~/.config/coyote:/home/agent/.config/coyote \
+  darkalex17/coyote
+```
+
+If you use the local vault provider and want your vault credentials available in the container, also mount the password file:
+
+```bash
+docker run --rm -it \
+  -v ~/.config/coyote:/home/agent/.config/coyote \
+  -v ~/.coyote_password:/home/agent/.coyote_password:ro \
+  darkalex17/coyote
 ```
 
 ### Scripts
 #### Linux/MacOS (`bash`)
-You can use the following command to run a bash script that downloads and installs the latest version of `loki` for your
+You can use the following command to run a bash script that downloads and installs the latest version of `coyote` for your
 OS (Linux/MacOS) and architecture (x86_64/arm64):
 
 ```shell
-curl -fsSL https://raw.githubusercontent.com/Dark-Alex-17/loki/main/install_loki.sh | bash
+curl -fsSL https://raw.githubusercontent.com/Dark-Alex-17/coyote/refs/heads/main/scripts/install_coyote.sh | bash
 ```
 
 #### Windows/Linux/MacOS (`PowerShell`)
-You can use the following command to run a PowerShell script that downloads and installs the latest version of `loki`
+You can use the following command to run a PowerShell script that downloads and installs the latest version of `coyote`
 for your OS (Windows/Linux/MacOS) and architecture (x86_64/arm64):
 
 ```powershell
-powershell -NoProfile -ExecutionPolicy Bypass -Command "iwr -useb https://raw.githubusercontent.com/Dark-Alex-17/loki/main/scripts/install_loki.ps1 | iex"
+powershell -NoProfile -ExecutionPolicy Bypass -Command "iwr -useb https://raw.githubusercontent.com/Dark-Alex-17/coyote/refs/heads/main/scripts/install_coyote.ps1 | iex"
 ```
 
 ### Manual
-Binaries are available on the [releases](https://github.com/Dark-Alex-17/loki/releases) page for the following platforms:
+Binaries are available on the [releases](https://github.com/Dark-Alex-17/coyote/releases) page for the following platforms:
 
 | Platform       | Architecture(s) |
 |----------------|-----------------|
@@ -118,102 +160,103 @@ Binaries are available on the [releases](https://github.com/Dark-Alex-17/loki/re
 #### Windows Instructions
 To use a binary from the releases page on Windows, do the following:
 
-1. Download the latest [binary](https://github.com/Dark-Alex-17/loki/releases) for your OS.
+1. Download the latest [binary](https://github.com/Dark-Alex-17/coyote/releases) for your OS.
 2. Use 7-Zip or TarTool to unpack the Tar file.
-3. Run the executable `loki.exe`!
+3. Run the executable `coyote.exe`!
 
 #### Linux/MacOS Instructions
 To use a binary from the releases page on Linux/MacOS, do the following:
 
-1. Download the latest [binary](https://github.com/Dark-Alex-17/loki/releases) for your OS.
+1. Download the latest [binary](https://github.com/Dark-Alex-17/coyote/releases) for your OS.
 2. `cd` to the directory where you downloaded the binary.
-3. Extract the binary with `tar -C /usr/local/bin -xzf loki-<arch>.tar.gz` (Note: This may require `sudo`)
-4. Now you can run `loki`!
+3. Extract the binary with `tar -C /usr/local/bin -xzf coyote-<arch>.tar.gz` (Note: This may require `sudo`)
+4. Now you can run `coyote`!
 
 ## Updating
-Loki can update itself in place to the latest GitHub release. Run `loki --update`
-for the newest release, or `loki --update v0.4.0` for a specific version:
+Coyote can update itself in place to the latest GitHub release. Run `coyote --update`
+for the newest release, or `coyote --update v0.4.0` for a specific version:
 
 ```shell
-loki --update
-loki --update v0.4.0
+coyote --update
+coyote --update v0.4.0
 ```
 
 The same is available from within the REPL via `.update` and `.update v0.4.0`.
 
-If Loki was installed with a package manager, prefer that package manager so its
-records stay in sync with the binary on disk; i.e. `brew upgrade loki` for Homebrew,
-or `cargo install --locked loki-ai` for Cargo.
+If Coyote was installed with a package manager, prefer that package manager so its
+records stay in sync with the binary on disk; i.e. `brew upgrade coyote` for Homebrew,
+or `cargo install --locked coyote-ai` for Cargo.
 
-When Loki detects a package-manager install it prints a warning and asks for
+When Coyote detects a package-manager install it prints a warning and asks for
 confirmation. In a non-interactive shell (no TTY), pass `--force` to update
 anyway:
 
 ```shell
-loki --update --force
+coyote --update --force
 ```
 
 ## Getting Started
 After installation, you can generate the configuration files and directories by simply running:
 
 ```sh
-loki --info
+coyote --info
 ```
 
-Then, you need to set up the Loki vault by creating a vault password file. Loki will do this for you automatically and
+Then, you need to set up the Coyote vault by creating a vault password file. Coyote will do this for you automatically and
 guide you through the process when you first attempt to access the vault. So, to get started, you can run:
 
 ```sh
-loki --list-secrets
+coyote --list-secrets
 ```
 
 ### Authentication
 Each client in your configuration needs authentication (with a few exceptions; e.g. ollama). Most clients use an API key
-(set via `api_key` in the config or through the [vault](https://github.com/Dark-Alex-17/loki/wiki/Vault)). For providers that support OAuth (e.g. Claude Pro/Max 
+(set via `api_key` in the config or through the [vault](https://github.com/Dark-Alex-17/coyote/wiki/Vault)). For providers that support OAuth (e.g. Claude Pro/Max 
 subscribers, Google Gemini), you can authenticate with your existing subscription instead:
 
 ```yaml
 # In your config.yaml
 clients:
   - type: claude
+    name: my-claude-oauth
     auth: oauth # Indicate you want to authenticate with OAuth instead of an API key
 ```
 
 ```sh
-loki --authenticate claude
+coyote --authenticate my-claude-oauth
 # Or via the REPL: .authenticate
 ```
 
-For full details, see the [authentication documentation](https://github.com/Dark-Alex-17/loki/wiki/Clients#authentication).
+For full details, see the [authentication documentation](https://github.com/Dark-Alex-17/coyote/wiki/Clients#authentication).
 
 ### Tab-Completions
-You can also enable tab completions to make using Loki easier. To do so, add the following to your shell profile:
+You can also enable tab completions to make using Coyote easier. To do so, add the following to your shell profile:
 ```shell
 # Bash
 # (add to: `~/.bashrc`)
-source <(COMPLETE=bash loki) 
+source <(COMPLETE=bash coyote) 
 
 # Zsh
 # (add to: `~/.zshrc`)
-source <(COMPLETE=zsh loki)
+source <(COMPLETE=zsh coyote)
 
 # Fish
 # (add to: `~/.config/fish/config.fish`)
-source <(COMPLETE=fish loki | psub)
+source <(COMPLETE=fish coyote | psub)
 
 # Elvish
 # (add to: `~/.elvish/rc.elv`)
-eval (E:COMPLETE=elvish loki | slurp)
+eval (E:COMPLETE=elvish coyote | slurp)
 
 # PowerShell
 # (add to: `$PROFILE`)
 $env:COMPLETE = "powershell"
-loki | Out-String | Invoke-Expression
+coyote | Out-String | Invoke-Expression
 ```
 
 ### Shell Integration
-You can integrate Loki's Shell Assistant into your shell for enhanced command-line assistance. Add the code in the
-corresponding [shell integration script](./scripts/shell-integration) to your shell. Then, you can invoke Loki to convert natural language to 
+You can integrate Coyote's Shell Assistant into your shell for enhanced command-line assistance. Add the code in the
+corresponding [shell integration script](./scripts/shell-integration) to your shell. Then, you can invoke Coyote to convert natural language to 
 shell commands by pressing `Alt-e`. For example:
 
 ```shell
@@ -223,18 +266,18 @@ find . -name "*.md"
 ```
 
 ## Configuration
-The location of the global Loki configuration varies between systems, so you can use the following command to find your
+The location of the global Coyote configuration varies between systems, so you can use the following command to find your
 `config.yaml` file:
 
 ```shell
-loki --info | grep 'config_file' | awk '{print $2}'
+coyote --info | grep 'config_file' | awk '{print $2}'
 ```
 
 The configuration file consists of a number of settings. To see a full example configuration file with every setting
 defined, refer to the [example configuration file](./config.example.yaml).
 
 ### Default LLM
-The following settings are available to configure the default LLM that is used when you start Loki, and its
+The following settings are available to configure the default LLM that is used when you start Coyote, and its
 hyperparameters:
 
 | Setting       | Description                                                                                                                                             |
@@ -244,34 +287,34 @@ hyperparameters:
 | `top_p`       | The default `top_p` hyperparameter value to use for all models, with a range of (0,1) (or (0,2) for some models); <br>Used unless explicitly overridden |
 
 ### CLI Behavior
-You can use the following settings to modify the behavior of Loki:
+You can use the following settings to modify the behavior of Coyote:
 
 | Setting       | Default Value | Description                                                                                                                         |
 |---------------|---------------|-------------------------------------------------------------------------------------------------------------------------------------|
 | `stream`      | `true`        | Controls whether to use stream-style APIs when querying for completions from LLM providers                                          |
 | `save`        | `true`        | Controls whether to save each query/response to every model to `messages.md` for posterity; Useful for debugging                    |
 | `keybindings` | `emacs`       | Specifies which keybinding schema to use; can either be `emacs` or `vi`                                                             |
-| `editor`      | `null`        | What text editor Loki should use to edit the input buffer or session (e.g. `vim`, `emacs`, `nano`, `hx`); <br>Defaults to `$EDITOR` |
+| `editor`      | `null`        | What text editor Coyote should use to edit the input buffer or session (e.g. `vim`, `emacs`, `nano`, `hx`); <br>Defaults to `$EDITOR` |
 | `wrap`        | `no`          | Controls whether text is wrapped (can be `no`, `auto`, or some `<max_width>`                                                        |
 | `wrap_code`   | `false`       | Enables or disables the wrapping of code blocks                                                                                     |
 
 ### Preludes
-Preludes let you define the default behavior for the different operating modes of Loki. The available settings are
+Preludes let you define the default behavior for the different operating modes of Coyote. The available settings are
 shown below:
 
 | Setting         | Description                                                                                                                                                                                                                                                                                                 |
 |-----------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
-| `repl_prelude`  | This setting lets you specify a default `session` or `role` to use when starting Loki in [REPL](https://github.com/Dark-Alex-17/loki/wiki/REPL) mode. <br>Values can be <ul><li>`role:<name>` to define a role</li><li>`session:<name>` to define a session</li><li>`<session>:<role>` to define both a session and a role to use</li></ul> |
-| `cmd_prelude`   | This setting lets you specify a default `session` or `role` to use when running one-off queries in Loki via the CLI. <br>Values can be <ul><li>`role:<name>` to define a role</li><li>`session:<name>` to define a session</li><li>`<session>:<role>` to define both a session and a role to use</li></ul>  |
+| `repl_prelude`  | This setting lets you specify a default `session` or `role` to use when starting Coyote in [REPL](https://github.com/Dark-Alex-17/coyote/wiki/REPL) mode. <br>Values can be <ul><li>`role:<name>` to define a role</li><li>`session:<name>` to define a session</li><li>`<session>:<role>` to define both a session and a role to use</li></ul> |
+| `cmd_prelude`   | This setting lets you specify a default `session` or `role` to use when running one-off queries in Coyote via the CLI. <br>Values can be <ul><li>`role:<name>` to define a role</li><li>`session:<name>` to define a session</li><li>`<session>:<role>` to define both a session and a role to use</li></ul>  |
 | `agent_session` | This setting is used to specify a default session that all agents should start into, unless otherwise specified in the agent configuration. (e.g. `temp`, `default`)                                                                                                                                        |
 
 ### Appearance
-The appearance of Loki can be modified using the following settings:
+The appearance of Coyote can be modified using the following settings:
 
 | Setting       | Default Value | Description                                          |
 |---------------|---------------|------------------------------------------------------|
 | `highlight`   | `true`        | This setting enables or disables syntax highlighting |
-| `light_theme` | `false`       | This setting toggles light mode in Loki              |
+| `light_theme` | `false`       | This setting toggles light mode in Coyote              |
 
 ### Miscellaneous Settings
 | Setting              | Default Value | Description                                                                                                      |
@@ -283,7 +326,7 @@ The appearance of Loki can be modified using the following settings:
 
 ## History
 
-Loki began as a fork of [AIChat CLI](https://github.com/sigoden/aichat) and has since evolved into an independent project.
+Coyote began as a fork of [AIChat CLI](https://github.com/sigoden/aichat) and has since evolved into an independent project.
 
 See [CREDITS.md](./CREDITS.md) for full attribution and background.
 
@@ -291,3 +334,16 @@ See [CREDITS.md](./CREDITS.md) for full attribution and background.
 
 ## Creator
 * [Alex Clarke](https://github.com/Dark-Alex-17)
+
+---
+
+## License
+
+Coyote is licensed under the [GNU Affero General Public License v3.0](./LICENSE)
+(AGPL-3.0-only).
+
+Coyote began as a fork of [AIChat](https://github.com/sigoden/aichat)
+(Copyright (c) sigoden), which is licensed under the MIT License. Substantial
+portions of Coyote are derived from AIChat and remain available under the MIT
+License, preserved in [LICENSE-MIT](./LICENSE-MIT). See [NOTICE](./NOTICE) and
+[CREDITS.md](./CREDITS.md) for details.

assets/agents/.shared/utils.sh

@@ -7,14 +7,14 @@ set -euo pipefail
 #######################
 
 # Cache file name for detected project info
-_LOKI_PROJECT_CACHE=".loki-project.json"
+_COYOTE_PROJECT_CACHE=".coyote-project.json"
 
 # Read cached project detection if valid
 # Usage: _read_project_cache "/path/to/project"
 # Returns: cached JSON on stdout (exit 0) or nothing (exit 1)
 _read_project_cache() {
   local dir="$1"
-  local cache_file="${dir}/${_LOKI_PROJECT_CACHE}"
+  local cache_file="${dir}/${_COYOTE_PROJECT_CACHE}"
 
   if [[ -f "${cache_file}" ]]; then
     local cached
@@ -32,7 +32,7 @@ _read_project_cache() {
 _write_project_cache() {
   local dir="$1"
   local json="$2"
-  local cache_file="${dir}/${_LOKI_PROJECT_CACHE}"
+  local cache_file="${dir}/${_COYOTE_PROJECT_CACHE}"
 
   echo "${json}" > "${cache_file}" 2>/dev/null || true
 }
@@ -40,15 +40,57 @@ _write_project_cache() {
 _detect_heuristic() {
   local dir="$1"
 
+  local runner="" runner_type="" runner_targets=""
+  if [[ -f "${dir}/Taskfile.yml" || -f "${dir}/Taskfile.yaml" || -f "${dir}/taskfile.yml" || -f "${dir}/taskfile.yaml" ]]; then
+    runner="task" runner_type="taskfile"
+    runner_targets=$( (cd "${dir}" && task --list-all 2>/dev/null | sed -n 's/^\* \([^:[:space:]]*\):.*/\1/p') || true)
+  elif [[ -f "${dir}/justfile" || -f "${dir}/Justfile" ]]; then
+    runner="just" runner_type="just"
+    runner_targets=$( (cd "${dir}" && just --summary 2>/dev/null | tr ' ' '\n') || true)
+  elif [[ -f "${dir}/Makefile" || -f "${dir}/makefile" || -f "${dir}/GNUmakefile" ]]; then
+    runner="make" runner_type="make"
+    local mk mkfiles=()
+    for mk in Makefile makefile GNUmakefile; do
+      [[ -f "${dir}/${mk}" ]] && mkfiles+=("${dir}/${mk}")
+    done
+    runner_targets=$(sed -n 's/^\([A-Za-z0-9_][A-Za-z0-9_.-]*\):\([^=].*\|\)$/\1/p' "${mkfiles[@]}" 2>/dev/null | sort -u || true)
+  fi
+  if [[ -n "${runner}" && -n "${runner_targets}" ]]; then
+    _pick_target() {
+      local c
+      for c in "$@"; do
+        if grep -qx "${c}" <<<"${runner_targets}"; then
+          echo "${runner} ${c}"
+          return 0
+        fi
+      done
+      echo ""
+    }
+    local r_build r_test r_check r_lint r_fmt
+    r_build=$(_pick_target build compile)
+    r_test=$(_pick_target test tests unit)
+    r_check=$(_pick_target check vet typecheck build)
+    r_lint=$(_pick_target lint fmt-check)
+    r_fmt=$(_pick_target fmt format)
+    if [[ -n "${r_build}${r_test}${r_check}${r_lint}${r_fmt}" ]]; then
+      echo "{\"type\":\"${runner_type}\",\"build\":\"${r_build}\",\"test\":\"${r_test}\",\"check\":\"${r_check}\",\"lint\":\"${r_lint}\",\"fmt\":\"${r_fmt}\"}"
+      return 0
+    fi
+  fi
+
   # Rust
   if [[ -f "${dir}/Cargo.toml" ]]; then
-    echo '{"type":"rust","build":"cargo build","test":"cargo test","check":"cargo check"}'
+    echo '{"type":"rust","build":"cargo build","test":"cargo test","check":"cargo check","lint":"cargo clippy --no-deps -- -D warnings","fmt":"cargo fmt"}'
     return 0
   fi
 
   # Go
   if [[ -f "${dir}/go.mod" ]]; then
-    echo '{"type":"go","build":"go build ./...","test":"go test ./...","check":"go vet ./..."}'
+    local go_lint=""
+    if compgen -G "${dir}/.golangci.*" &>/dev/null && command -v golangci-lint &>/dev/null; then
+      go_lint="golangci-lint run"
+    fi
+    echo "{\"type\":\"go\",\"build\":\"go build ./...\",\"test\":\"go test ./...\",\"check\":\"go vet ./...\",\"lint\":\"${go_lint}\",\"fmt\":\"gofmt -w .\"}"
     return 0
   fi
 
@@ -65,7 +107,25 @@ _detect_heuristic() {
     [[ -f "${dir}/pnpm-lock.yaml" ]] && pm="pnpm"
     [[ -f "${dir}/yarn.lock" ]] && pm="yarn"
 
-    echo "{\"type\":\"nodejs\",\"build\":\"${pm} run build\",\"test\":\"${pm} test\",\"check\":\"${pm} run lint\"}"
+    # Emit only scripts the manifest actually declares (same introspection
+    # contract as the runner tier: never guess a target into existence).
+    _pkg_script() {
+      local s
+      for s in "$@"; do
+        if jq -e --arg s "$s" '.scripts[$s] // empty' "${dir}/package.json" &>/dev/null; then
+          echo "${pm} run ${s}"
+          return 0
+        fi
+      done
+      echo ""
+    }
+    local p_build p_test p_check p_lint p_fmt
+    p_build=$(_pkg_script build compile)
+    p_test=$(_pkg_script test)
+    p_check=$(_pkg_script check typecheck tsc)
+    p_lint=$(_pkg_script lint)
+    p_fmt=$(_pkg_script fmt format prettier)
+    echo "{\"type\":\"nodejs\",\"build\":\"${p_build}\",\"test\":\"${p_test}\",\"check\":\"${p_check}\",\"lint\":\"${p_lint}\",\"fmt\":\"${p_fmt}\"}"
     return 0
   fi
 
@@ -82,7 +142,7 @@ _detect_heuristic() {
       check_cmd="uv run ruff check ."
     fi
 
-    echo "{\"type\":\"python\",\"build\":\"\",\"test\":\"${test_cmd}\",\"check\":\"${check_cmd}\"}"
+    echo "{\"type\":\"python\",\"build\":\"\",\"test\":\"${test_cmd}\",\"check\":\"${check_cmd}\",\"lint\":\"${check_cmd}\",\"fmt\":\"ruff format .\"}"
     return 0
   fi
 
@@ -144,17 +204,6 @@ _detect_heuristic() {
     return 0
   fi
 
-  # Generic build systems (last resort before LLM)
-  if [[ -f "${dir}/justfile" ]] || [[ -f "${dir}/Justfile" ]]; then
-    echo '{"type":"just","build":"just build","test":"just test","check":"just lint"}'
-    return 0
-  fi
-
-  if [[ -f "${dir}/Makefile" ]] || [[ -f "${dir}/makefile" ]] || [[ -f "${dir}/GNUmakefile" ]]; then
-    echo '{"type":"make","build":"make build","test":"make test","check":"make lint"}'
-    return 0
-  fi
-
   return 1
 }
 
@@ -218,7 +267,9 @@ _detect_with_llm() {
   local prompt
   prompt=$(cat <<-EOF
 
-		Analyze this project directory and determine the project type, primary language, and the correct shell commands to build, test, and check (lint/typecheck) it.
+		Analyze this project directory and determine the project type, primary language, and the correct shell commands to build, test, check (typecheck/vet), lint, and format it.
+
+		PRIORITY RULE: if the project declares its own task-runner interface (a Taskfile, justfile, Makefile, package.json scripts, or similar), those declared targets ARE the correct commands — prefer them over generic ecosystem defaults, and never invent a target the interface does not declare.
 
 		EOF
 	)
@@ -226,25 +277,25 @@ _detect_with_llm() {
   prompt+=$(cat <<-EOF
 
 		Respond with ONLY a valid JSON object. No markdown fences, no explanation, no extra text.
-		The JSON must have exactly these 4 keys:
-		{"type":"<language>","build":"<build command>","test":"<test command>","check":"<lint or typecheck command>"}
+		The JSON must have exactly these 6 keys:
+		{"type":"<language>","build":"<build command>","test":"<test command>","check":"<typecheck/vet command>","lint":"<lint command>","fmt":"<format command>"}
 
 		Rules:
 		- "type" must be a single lowercase word (e.g. rust, go, python, nodejs, java, ruby, elixir, cpp, c, zig, haskell, scala, kotlin, dart, swift, php, dotnet, etc.)
-		- If a command doesn't apply to this project, use an empty string, ""
+		- If a command doesn't apply to this project, use an empty string, "" — NEVER guess a command that might not exist; a wrongly-guessed command is worse than an empty one
 		- Use the most standard/common commands for the detected ecosystem
 		- If you detect a package manager lockfile, use that package manager (e.g. pnpm over npm)
 		EOF
 	)
 
   local llm_response
-  llm_response=$(loki --no-stream "${prompt}" 2>/dev/null) || return 1
+  llm_response=$(coyote --no-stream "${prompt}" 2>/dev/null) || return 1
 
   llm_response=$(echo "${llm_response}" | sed 's/^```json//;s/^```//;s/```$//' | tr -d '\n' | sed 's/^[[:space:]]*//')
   llm_response=$(echo "${llm_response}" | grep -o '{[^}]*}' | head -1)
 
   if echo "${llm_response}" | jq -e '.type and .build != null and .test != null and .check != null' &>/dev/null; then
-    echo "${llm_response}" | jq -c '{type: (.type // "unknown"), build: (.build // ""), test: (.test // ""), check: (.check // "")}'
+    echo "${llm_response}" | jq -c '{type: (.type // "unknown"), build: (.build // ""), test: (.test // ""), check: (.check // ""), lint: (.lint // ""), fmt: (.fmt // "")}'
     return 0
   fi
 
@@ -258,7 +309,7 @@ detect_project() {
 
   local cached
   if cached=$(_read_project_cache "${dir}"); then
-    echo "${cached}" | jq -c '{type, build, test, check}'
+    echo "${cached}" | jq -c '{type, build, test, check, lint: (.lint // ""), fmt: (.fmt // "")}'
     return 0
   fi
 
@@ -286,6 +337,31 @@ detect_project() {
   echo '{"type":"unknown","build":"","test":"","check":""}'
 }
 
+# resolve_gate_dir maps a workspace root to the directory verification gates
+# must run in. A delivery-repo worker's workspace root holds only dotfiles
+# plus the clone, so gates aimed at the root detect nothing and silently
+# no-op. When the root has no project markers and exactly ONE first-level
+# git repo exists, gates run inside it; anything ambiguous stays at the root.
+resolve_gate_dir() {
+  local dir="${1:-.}"
+  local m
+  for m in Taskfile.yml Taskfile.yaml taskfile.yml Cargo.toml go.mod package.json pyproject.toml setup.py pom.xml build.gradle mix.exs Gemfile composer.json Makefile justfile Justfile CMakeLists.txt; do
+    if [[ -e "${dir}/${m}" ]]; then
+      echo "${dir}"
+      return 0
+    fi
+  done
+  local repos=() d
+  for d in "${dir}"/*/; do
+    [[ -d "${d}/.git" ]] && repos+=("${d}")
+  done
+  if [[ ${#repos[@]} -eq 1 ]]; then
+    echo "${repos[0]%/}"
+    return 0
+  fi
+  echo "${dir}"
+}
+
 ###########################
 ## FILE SEARCH UTILITIES ##
 ###########################

assets/agents/adversary/README.md

@@ -0,0 +1,94 @@
+# Adversary
+
+An **adversarial plan-conformance reviewer**. Where [`code-reviewer`](../code-reviewer/README.md)
+asks *"is this code good?"*, `adversary` asks a different, harder question:
+
+> **"Is this the code the plan asked for — all of it, and only it?"**
+
+It hunts the gap between what a task/plan *specified* and what the implementer actually *built*:
+silently skipped acceptance criteria, scope creep, interface substitution, approach drift, and the
+requirements that never showed up in the diff at all ("the dog that didn't bark"). It assumes the
+implementer drifted until the diff proves otherwise — the independence is the value.
+
+## Why it's separate from `code-reviewer`
+
+| | `code-reviewer` | `adversary` |
+|---|---|---|
+| Question | Is the code correct/clean/safe? | Does the code match the plan? |
+| Input | The diff | The diff **+ the plan's acceptance criteria** |
+| Blind spot it covers | slop, bugs, coupling, footguns | skipped criteria, scope drift, contract breakage |
+| Output | severity-tagged findings (🔴🟡🟢) | a blocking verdict: `CONFORMS` / `DIVERGES` |
+
+They are **complementary passes**, not substitutes. `sisyphus` runs both on non-trivial work: one
+guards quality, the other guards fidelity to the plan.
+
+## Verdict (blocking)
+
+The agent ends every review with one sentinel:
+
+```
+ADVERSARIAL_REVIEW: CONFORMS
+Criteria: N/N met (all with tests).
+```
+
+```
+ADVERSARIAL_REVIEW: DIVERGES
+Criteria: X/N met, Y partial, Z unmet/diverged.
+Complaints:
+1. Acceptance criterion "<quoted>" — <Unmet|Partial|Diverged> — <what the diff does/omits, file:line> — <fix>
+2. ...
+```
+
+A `DIVERGES` verdict **blocks** completion. The caller (sisyphus/architect) must reconcile it —
+resume the SAME coder/sisyphus session with the complaints pasted verbatim — or escalate. It mirrors
+the `oracle` + `plan-review` gate used before implementation, but applied *after* implementation.
+
+Every complaint ties to a quoted acceptance criterion (or a named scope/interface/out-of-scope
+violation) and cites `file:line`. Vague complaints are not emitted.
+
+## How it reviews
+
+Driven by the [`adversarial-review`](../../skills/adversarial-review/SKILL.md) skill:
+
+1. Map **every** acceptance criterion to specific evidence in the diff → ✅ Met / ⚠️ Partial / ❌ Unmet / 🔀 Diverged. No test proving the behavior ⇒ at best ⚠️ Partial.
+2. Ground-truth with read-only tools (`fs_grep`/`fs_read`/`ast_grep`): confirm required symbols exist as specified, changes land where they must, new behavior is actually reached, tests target behavior not implementation.
+3. Hunt adversarially for the **absent**: skipped criteria, scope creep, interface/approach substitution, out-of-scope touches, downstream contract breakage.
+
+It is **read-only** — it produces a verdict, never a fix.
+
+## Usage
+
+Typically spawned by `sisyphus` (or `architect`) alongside `code-reviewer`. The spawn prompt IS its
+entire context, so it must include the diff (or a base ref to fetch) **and** the acceptance criteria:
+
+```sh
+agent__spawn --agent adversary --prompt "
+## TASK
+Adversarially review the recent changes for TASK-NNN against its plan. Return CONFORMS/DIVERGES.
+
+## DIFF
+Run get_diff (or --base main), or: <paste diff>
+
+## PLAN — acceptance criteria to check against
+<paste the task index.md body + the relevant PLAN-*.md section, verbatim>
+"
+```
+
+Direct invocation for ad-hoc use:
+
+```sh
+coyote -a adversary --agent-variable project_dir /path/to/repo \
+  "Review staged changes against these criteria: <paste criteria>"
+```
+
+### Tools
+
+- `get_diff [--base <ref>]` — staged → unstaged → `HEAD~1` fallback (or an explicit base/PR branch).
+- `get_changed_files [--base <ref>]` — quick changed-file map.
+- Plus read-only `fs_*` and `ast_grep` for ground-truth checks.
+
+## Related
+
+- [`adversarial-review`](../../skills/adversarial-review/SKILL.md) — the conformance methodology it runs on.
+- [`code-reviewer`](../code-reviewer/README.md) — the quality reviewer it runs alongside.
+- [`plan-review`](../../skills/plan-review/SKILL.md) — the *pre*-implementation plan gate; `adversary` is its *post*-implementation counterpart.

assets/agents/adversary/config.yaml

@@ -0,0 +1,118 @@
+name: adversary
+description: Adversarial plan-conformance reviewer - judges whether an implementation matches the task/plan it was supposed to satisfy (not code quality). Returns a blocking CONFORMS/DIVERGES verdict. Complements code-reviewer. Designed to be delegated to by sisyphus.
+version: 1.0.0
+
+auto_continue: true
+max_auto_continues: 15
+inject_todo_instructions: true
+
+skills_enabled: true
+enabled_skills:
+  - adversarial-review
+
+variables:
+  - name: project_dir
+    description: Project directory containing the changes under review
+    default: '.'
+  - name: auto_confirm
+    description: Auto-confirm command execution
+    default: '1'
+
+global_tools:
+  - ast_grep.sh
+  - fs_read.sh
+  - fs_cat.sh
+  - fs_grep.sh
+  - fs_glob.sh
+  - fs_ls.sh
+  - execute_command.sh
+
+instructions: |
+  You are an adversarial plan-conformance reviewer. You answer ONE question: **does this
+  implementation match the plan it was supposed to satisfy — all of it, and only it?** You are NOT
+  the code-quality reviewer (that is `code-reviewer`/`file-reviewer`, which judges correctness, slop,
+  and style). You judge CONFORMANCE: skipped acceptance criteria, silent scope drift, interface
+  substitution, and things the plan required that never showed up in the diff.
+
+  Your value is independence and suspicion. Assume the implementer drifted, cut a corner, or misread
+  the plan until the diff proves otherwise.
+
+  ## Step 0: Load the skill
+
+  Before anything else, `skill__load` `adversarial-review`. It carries your methodology: the
+  criterion-by-criterion evidence mapping, the adversarial checklist (silently skipped criteria,
+  scope drift, interface drift, ground-truth verification, out-of-scope violations, downstream
+  contract breakage), and the exact verdict format. The skill body is your source of truth for HOW to
+  review and WHAT to flag; these instructions handle workflow and I/O.
+
+  ## Input (the spawn prompt IS your entire context)
+
+  You are given:
+  1. **The diff** — pasted inline, or run `get_diff` (optionally `--base <ref>`) if told to fetch it.
+  2. **The plan** — the task's Objective, Tasks, and especially its **Acceptance criteria**, pasted
+     inline (e.g. a task file's What/Steps/Acceptance criteria + the relevant plan section), or a path to read.
+
+  If the plan / acceptance criteria are missing, STOP and say so: conformance cannot be judged
+  without a spec. Do not invent criteria or guess intent.
+
+  ## Workflow
+
+  1. Load `adversarial-review`.
+  2. Get the diff (inline or via `get_diff`) and identify the changed files.
+  3. For EACH acceptance criterion: find the specific evidence in the diff that satisfies it and
+     classify it ✅ Met / ⚠️ Partial / ❌ Unmet / 🔀 Diverged. A criterion with no test proving its
+     behavior is at best ⚠️ Partial.
+  4. Ground-truth every claim: `fs_grep` the symbols the plan requires (confirm they exist, spelled
+     as specified), `fs_read` around each hunk to confirm the change makes the criterion true, grep
+     callers to confirm new behavior is reached, confirm tests target behavior not implementation.
+     Use `ast_grep` for structural checks (e.g. "was this function signature actually changed?").
+  5. Hunt adversarially for what's ABSENT (the dog that didn't bark), scope creep, interface/approach
+     substitution, out-of-scope touches, and downstream contract breakage — per the skill checklist.
+  6. Emit the verdict in the skill's exact format.
+
+  ## Output — verdict (MANDATORY, exact format)
+
+  End with EXACTLY one of these sentinels so the caller can route on it:
+
+  ```
+  ADVERSARIAL_REVIEW: CONFORMS
+  Criteria: N/N met (all with tests).
+  <optional: 1-3 non-blocking observations>
+  ```
+
+  ```
+  ADVERSARIAL_REVIEW: DIVERGES
+  Criteria: X/N met, Y partial, Z unmet/diverged.
+  Complaints:
+  1. Acceptance criterion "<quoted>" — <Unmet|Partial|Diverged> — <what the diff does/omits, file:line> — <what would make it conform>
+  2. Scope drift / interface drift / out-of-scope — <file:line> — <the violation> — <the fix>
+  3. ...
+  ```
+
+  Every complaint MUST quote the specific acceptance criterion (or name the specific scope/interface/
+  out-of-scope violation) AND cite file:line. A complaint with no criterion reference and no location
+  is noise — do not emit it.
+
+  ## Rules
+
+  1. **You are read-only.** Never modify files. You produce a verdict; the implementer owns the fix.
+  2. **Conformance, not quality.** Do not flag style/naming/micro-optimizations unless they cause a
+     criterion to be unmet. If a quality defect breaks a criterion (a race violating a correctness
+     criterion), flag it as a conformance failure and note it is also a quality issue.
+  3. **No test ⇒ not met.** An acceptance criterion is a promise of observable behavior; unproven
+     behavior is at best Partial.
+  4. **Absence is a finding.** Review what SHOULD be in the diff per the plan, not only what IS.
+  5. **Don't re-litigate a settled decision** — but DO flag when the diff silently overrode one the
+     plan recorded ("do X not Y because Z" → diff does Y).
+  6. **The plan can be the culprit.** If the plan is impossible/self-contradictory, that is DIVERGES
+     with the plan named as root cause — never judge against a plan you silently corrected.
+  7. Be terse and decisive. Three real divergences beat fifteen weak ones. If everything is a nitpick,
+     it CONFORMS — say so.
+
+  ## Context
+  - Project: {{project_dir}}
+  - CWD: {{__cwd__}}
+  - Shell: {{__shell__}}
+
+  ## Available Tools
+  {{__tools__}}

assets/agents/adversary/tools.sh

@@ -0,0 +1,78 @@
+#!/usr/bin/env bash
+set -eo pipefail
+
+# @env LLM_OUTPUT=/dev/stdout
+# @env LLM_AGENT_VAR_PROJECT_DIR=.
+# @describe Adversarial plan-conformance reviewer tools
+
+_project_dir() {
+  local dir="${LLM_AGENT_VAR_PROJECT_DIR:-.}"
+  (cd "${dir}" 2>/dev/null && pwd) || echo "${dir}"
+}
+
+# @cmd Get the git diff to review for plan conformance. Returns staged changes, or unstaged if nothing is staged, or the HEAD~1 diff if the working tree is clean.
+# @option --base Optional base ref to diff against (e.g., "main", "HEAD~3", a commit SHA, or a PR base branch)
+get_diff() {
+  local project_dir
+  project_dir=$(_project_dir)
+  # shellcheck disable=SC2154
+  local base="${argc_base:-}"
+
+  local diff_output=""
+  if [[ -n "${base}" ]]; then
+    diff_output=$(cd "${project_dir}" && git diff "${base}" 2>&1) || true
+  else
+    diff_output=$(cd "${project_dir}" && git diff --cached 2>&1) || true
+    if [[ -z "${diff_output}" ]]; then
+      diff_output=$(cd "${project_dir}" && git diff 2>&1) || true
+    fi
+    if [[ -z "${diff_output}" ]]; then
+      diff_output=$(cd "${project_dir}" && git diff HEAD~1 2>&1) || true
+    fi
+  fi
+
+  if [[ -z "${diff_output}" ]]; then
+    echo "No changes found to review in ${project_dir}." >> "$LLM_OUTPUT"
+    return 0
+  fi
+
+  local file_count
+  file_count=$(echo "${diff_output}" | grep -c '^diff --git' || true)
+  {
+    echo "Diff contains changes to ${file_count} file(s):"
+    echo ""
+    echo "${diff_output}"
+  } >> "$LLM_OUTPUT"
+}
+
+# @cmd Get the list of changed files with stats (a quick map of what to check against the plan).
+# @option --base Optional base ref to diff against
+get_changed_files() {
+  local project_dir
+  project_dir=$(_project_dir)
+  local base="${argc_base:-}"
+
+  local stat_output=""
+  if [[ -n "${base}" ]]; then
+    stat_output=$(cd "${project_dir}" && git diff --stat "${base}" 2>&1) || true
+  else
+    stat_output=$(cd "${project_dir}" && git diff --cached --stat 2>&1) || true
+    if [[ -z "${stat_output}" ]]; then
+      stat_output=$(cd "${project_dir}" && git diff --stat 2>&1) || true
+    fi
+    if [[ -z "${stat_output}" ]]; then
+      stat_output=$(cd "${project_dir}" && git diff --stat HEAD~1 2>&1) || true
+    fi
+  fi
+
+  if [[ -z "${stat_output}" ]]; then
+    echo "No changes found in ${project_dir}." >> "$LLM_OUTPUT"
+    return 0
+  fi
+
+  {
+    echo "Changed files:"
+    echo ""
+    echo "${stat_output}"
+  } >> "$LLM_OUTPUT"
+}

assets/agents/architect/README.md

@@ -0,0 +1,182 @@
+# Architect
+
+A **design-doc orchestrator for any project**. Give it one high-level design doc; it decomposes the
+doc into a quality-gated plan and ~1-engineer-day task files, spawns **one
+[Sisyphus](../sisyphus/README.md) per task** on a single run branch, verifies each task with an
+adversarial plan-conformance check, and finishes with **one draft PR** (CI checks watched to green)
+plus tracked follow-up tasks for the manual work the code can't do for itself.
+
+Architect does **not** write feature code itself. It owns the *process*; Sisyphus owns each *task*.
+
+## The pipeline it drives
+
+```mermaid
+flowchart TD
+    user([Design doc]) --> architect["Architect<br/>design-doc orchestrator"]
+    architect --> orient["Phase A — Orient<br/>project conventions · build/test commands · design doc"]
+    orient --> design["Phase B — design-session<br/>plans_dir/PLAN-&lt;slug&gt;.md + 1-day task breakdown"]
+
+    design -. "grounding" .-> explore[["explore<br/>codebase grep<br/>× parallel"]]
+    design -. "unfamiliar libraries" .-> librarian[["librarian<br/>docs + OSS grep"]]
+    explore -. "findings ground<br/>the breakdown" .-> design
+    librarian -. "findings ground<br/>the breakdown" .-> design
+
+    design --> gatekeeper[["gatekeeper<br/>self-containedness audit<br/>(docker-container test)"]]
+    gatekeeper --> g1{"PLAN_GATE?"}
+    g1 -->|"LEAKY (≤ 2 cycles)"| amend["Answer the missing questions<br/>via explore / librarian / docs<br/>(user__ask only for business rules)<br/>→ amend the plan"]
+    amend --> gatekeeper
+    g1 -->|"LEAKY after 2 cycles"| escalate
+    g1 -->|"SEALED"| oracle[["oracle<br/>plan-review<br/>(executability)"]]
+
+    oracle --> g2{"PLAN_REVIEW?"}
+    g2 -->|"REJECT — fix complaints,<br/>re-submit SAME session"| oracle
+    g2 -->|"OKAY"| tasks["Phase D — materialize tasks<br/>plans_dir/tasks/TASK-NNN-*/ (task-tracking)"]
+
+    tasks --> branch["Phase E — run branch<br/>feat/PLAN-&lt;slug&gt; off base_branch"]
+    branch --> claim["Claim task (sequential, dependency order)<br/>status: in-progress + base SHA"]
+    claim --> sisyphus[["sisyphus<br/>implement ONE task on the run branch<br/>commit + push — NO PR"]]
+    sisyphus --> adversary[["adversary<br/>conformance check<br/>diff vs task base SHA"]]
+
+    adversary --> verdict{"ADVERSARIAL_REVIEW?"}
+    verdict -->|"DIVERGES — resume<br/>SAME sisyphus session (once)"| sisyphus
+    verdict -->|"still DIVERGES"| escalate
+    verdict -->|"CONFORMS"| taskdone["Close task<br/>status: complete · log commits + follow-ups"]
+
+    taskdone --> more{"More tasks?"}
+    more -->|"yes"| claim
+    more -->|"no"| finish["Phase F — full build + tests<br/>on the integrated run branch"]
+
+    finish --> pr["ONE DRAFT PR: run branch → base_branch<br/>(never marked ready — user reviews first)<br/>body: task checklist + Follow-up / manual actions"]
+    pr --> checks{"PR runs/checks<br/>green?"}
+    checks -->|"failure — resume responsible<br/>sisyphus session, fix, push"| checks
+    checks -->|"external flake /<br/>broken base branch"| escalate
+    checks -->|"green"| followups["Create follow-up task files<br/>(type: followup, pending)<br/>→ picked up by the user post-merge"]
+    followups --> backfill["Backfill PR link into PLAN + task logs<br/>PLAN status: implemented"]
+    backfill --> validate["task-tracking consistency checks"]
+    validate --> done([Run complete])
+
+    escalate([user__ask — escalate to user])
+
+    branch -. "parallel_tasks=1 (opt-in):<br/>per-task worktrees + task branches,<br/>merged one at a time with<br/>integration tests after every merge" .-> claim
+```
+
+## Where state lives
+
+Everything is file-based in **`plans_dir`** (default `plans/`, resolved against the project):
+
+```
+<plans_dir>/
+  PLAN-<slug>.md                  # problem / approach / alternatives / task breakdown
+  tasks/TASK-NNN-<slug>/
+    index.md                      # What / Steps / Acceptance criteria; status in frontmatter
+    log.md                        # append-only audit trail (branch, commits, follow-ups, PR)
+```
+
+- `plans_dir` **inside the repo** (default) → planning files ride the run branch and land in the PR
+  (self-documenting review).
+- `plans_dir` **absolute, outside the repo** (e.g. a common runs directory) → nothing planning-related
+  is ever committed.
+
+Disk is the durable store: task statuses, logs, and follow-ups survive context compression; chat
+history does not.
+
+## The three review gates
+
+| Gate | Agent | Question | When |
+|------|-------|----------|------|
+| Self-containedness | [`gatekeeper`](../gatekeeper/README.md) | "Can a context-free LLM implement from this plan alone?" | Before tasks exist |
+| Executability | `oracle` + `plan-review` | "Is the approach sound, verifiable, correctly ordered?" | After sealing |
+| Conformance | [`adversary`](../adversary/README.md) | "Is the built code what the plan asked for?" | After each task |
+
+## Key conventions it enforces
+
+- **One task = one engineer-day** — anything larger gets decomposed at the design stage.
+- **Task state on disk** — `status:` frontmatter lifecycle per the `task-tracking` skill; no state
+  lives only in chat.
+- **One run branch, one draft PR** — `feat/PLAN-<slug>` off `base_branch`; the PR is never opened
+  per-task, never non-draft, never marked ready-for-review (you flip it yourself).
+- **CI checks watched to green** — failures are routed back to the responsible Sisyphus session; the
+  run isn't done with red or pending checks.
+- **No plan references in code comments** — comments never cite the design doc, plan, phases, steps,
+  or TASK numbers (docs drift; comments rot). Plan references live in commit messages only.
+- **`.env` never lands in a repo** — only `.env.example` with placeholder keys; real values become a
+  follow-up.
+- **Follow-ups are tracked, never dropped** — every manual action (secrets, cloud roles, console
+  steps, cross-repo changes) is reported per task, logged durably, rolled into the PR's
+  `## Follow-up / manual actions` section (pre-merge items first), and materialized as
+  `type: followup` task files for you to pick up post-merge.
+
+## Usage
+
+```sh
+# From the target project root (default autonomy: full)
+coyote -a architect --agent-variable design_doc docs/design/my-feature.md \
+  "Implement this design doc end to end"
+
+# Approve the task breakdown once, then run autonomously
+coyote -a architect \
+  --agent-variable design_doc docs/design/my-feature.md \
+  --agent-variable autonomy plan-gate \
+  "Decompose and implement"
+
+# Different project / plans outside the repo / PR against a non-main base
+coyote -a architect \
+  --agent-variable project_dir ~/code/my-service \
+  --agent-variable plans_dir ~/architect-runs/my-service \
+  --agent-variable base_branch develop \
+  --agent-variable design_doc ~/docs/big-refactor.md \
+  "Run the pipeline"
+```
+
+### Variables
+
+| Variable | Default | Meaning |
+|----------|---------|---------|
+| `project_dir` | `.` | The target repo — the only WRITE target for feature code. |
+| `plans_dir` | `plans` | Where PLAN + task files live. Relative → in-repo (rides the PR); absolute → outside git. |
+| `design_doc` | *(empty)* | Path to the design doc; asked for if unset. |
+| `base_branch` | `main` | Branch the run branch forks from and the PR targets. |
+| `autonomy` | `full` | `full` (no gates) · `plan-gate` (approve breakdown once) · `phase-gate` (approve each task). |
+| `parallel_tasks` | `0` | `0` = sequential (default) · `1` = opt-in worktree-parallel execution for eligible tasks. |
+| `auto_confirm` | `1` | Skip the shell confirm guard (needed for non-interactive autonomous runs). |
+
+## Autonomy
+
+Fully autonomous end-to-end by default — it halts only for genuine blockers: scope-changing
+ambiguity or unresolved design questions, a task that fails after Sisyphus's own recovery (consults
+Oracle, then escalates), and any destructive/irreversible action. Use `plan-gate` or `phase-gate`
+to insert approval checkpoints.
+
+## Parallel task execution (opt-in)
+
+By default (`parallel_tasks: 0`) tasks run **sequentially** on the single run branch. Setting
+`parallel_tasks: 1` enables worktree-based parallelism:
+
+- Eligible tasks (mutually unblocked, plan-declared file-disjoint, max 3 concurrent) each get an
+  isolated `git worktree` + task branch forked from the run branch tip.
+- Tasks touching **migrations, generated code, or dependency manifests/lockfiles** are never
+  parallel-eligible — shared hotspots collide even when the plan calls tasks independent.
+- Architect integrates: completed task branches merge into the run branch **one at a time**, with a
+  full build + test run after every merge. Conflicts go back to that task's Sisyphus session to
+  rebase and re-verify.
+- Worktrees and task branches are cleaned up after each clean merge. Phase F (single draft PR +
+  CI-check watch) is unchanged in both modes.
+
+## Sub-agents it spawns
+
+| Agent | Used for |
+|-------|----------|
+| [`sisyphus`](../sisyphus/README.md) | Implement ONE task's code (its own explore→coder→verify→review loop). One per task. |
+| [`gatekeeper`](../gatekeeper/README.md) | Plan self-containedness gate (`PLAN_GATE: SEALED/LEAKY`). |
+| [`adversary`](../adversary/README.md) | Per-task plan-conformance verdict (`ADVERSARIAL_REVIEW: CONFORMS/DIVERGES`). |
+| [`oracle`](../oracle/README.md) | Plan review (`plan-review`); diagnosis when a task fails after Sisyphus recovery. |
+| [`explore`](../explore/README.md) | Ground the design/plan in real code; read other local repos for library usage and call sites. |
+| [`librarian`](../librarian/README.md) | External docs / OSS examples for unfamiliar libraries. |
+
+## Related skills
+
+- [`design-session`](../../skills/design-session/SKILL.md) — design doc → grounded proposal → PLAN + sized breakdown.
+- [`task-tracking`](../../skills/task-tracking/SKILL.md) — the task-file schema, lifecycle, and consistency checks.
+- [`plan-gatekeeping`](../../skills/plan-gatekeeping/SKILL.md) — the gatekeeper's self-containedness manifest.
+- [`plan-authoring`](../../skills/plan-authoring/SKILL.md) / [`plan-review`](../../skills/plan-review/SKILL.md) — plan schema + oracle's executability review.
+- [`adversarial-review`](../../skills/adversarial-review/SKILL.md) — the adversary's conformance methodology.

assets/agents/architect/config.yaml

@@ -0,0 +1,460 @@
+name: architect
+description: |
+  Design-doc orchestrator for any project. Consumes a high-level design doc, decomposes it into a
+  gated plan (gatekeeper self-containedness + oracle plan-review) and ~1-engineer-day task files,
+  spawns one Sisyphus per task on a single run branch, verifies each with an adversarial
+  plan-conformance check, and finishes with ONE draft PR (CI checks watched to green) plus tracked
+  follow-up tasks. Task state lives on disk in a plans directory, so runs survive context compression.
+version: 2.0.0
+agent_session: temp
+auto_continue: true
+max_auto_continues: 100
+inject_todo_instructions: true
+
+can_spawn_agents: true
+spawnable_agents:
+  - sisyphus
+  - oracle
+  - explore
+  - librarian
+  - adversary
+  - gatekeeper
+max_concurrent_agents: 10
+max_agent_depth: 10
+inject_spawn_instructions: true
+summarization_threshold: 100000
+
+skills_enabled: true
+enabled_skills:
+  - design-session
+  - task-tracking
+  - plan-authoring
+  - delegation-protocol
+  - git-master
+  - parallel-research
+
+variables:
+  - name: project_dir
+    description: Absolute path to the target project repo — the ONLY write target for feature code
+    default: '.'
+  - name: plans_dir
+    description: Where the PLAN file and task dirs live. Relative paths resolve against project_dir (and then ride the run branch into the PR); an absolute path outside the repo keeps planning files out of git entirely.
+    default: 'plans'
+  - name: design_doc
+    description: Path to the high-level design doc to implement (absolute, or relative to project_dir)
+    default: ''
+  - name: base_branch
+    description: The branch the run branch forks from and the PR targets
+    default: 'main'
+  - name: autonomy
+    description: 'How autonomous the run is: full (no gates), plan-gate (approve breakdown once, then autonomous), phase-gate (approve each task)'
+    default: full
+  - name: auto_confirm
+    description: Auto-confirm command execution (1 = skip the shell guard_operation TTY prompt, needed for non-interactive autonomous runs)
+    default: '1'
+  - name: parallel_tasks
+    description: 'Opt-in worktree-based parallel task execution: 0 = sequential (default, one task at a time on the run branch), 1 = eligible tasks run as concurrent Sisyphus agents in isolated git worktrees, merged back one at a time'
+    default: '0'
+
+global_tools:
+  - ast_grep.sh
+  - fs_read.sh
+  - fs_grep.sh
+  - fs_glob.sh
+  - fs_ls.sh
+  - fs_write.sh
+  - fs_patch.sh
+  - fs_mkdir.sh
+  - execute_command.sh
+
+instructions: |
+  You are **Architect** — an orchestrator that takes a single high-level design doc and drives it
+  end-to-end to implementation on ANY project. You do NOT write feature code yourself. You decompose,
+  gate the plan, delegate one task to one **Sisyphus** sub-agent, verify conformance, track state on
+  disk, and finish with a single draft PR — repeating until the entire design doc is implemented.
+
+  ## Ground rules — READ BEFORE ANYTHING
+
+  **Write target.** ALL feature code goes in {{project_dir}}. You and your sub-agents MAY freely READ
+  other local repos/directories (internal libraries, legacy patterns, call sites, shared contracts)
+  — reading is encouraged; WRITING anywhere but {{project_dir}} is a scope violation. If the design
+  genuinely requires writing outside {{project_dir}}, STOP and escalate; likely it's a follow-up.
+
+  **Git model — one run branch, one draft PR.** All work lands on a single RUN BRANCH
+  (`feat/PLAN-<slug>`, forked from {{base_branch}}), and exactly ONE DRAFT PR is opened at the END of
+  the run (Phase F) covering the entire design doc — NEVER one PR per task, NEVER a push to
+  {{base_branch}}. Before any `git push`/branch/PR, confirm you are in {{project_dir}}
+  (`git remote get-url origin`).
+
+  **Task state lives on disk.** {{plans_dir}} (relative → resolved against {{project_dir}}, riding
+  the run branch into the PR; absolute → outside git entirely) holds `PLAN-<slug>.md` and
+  `tasks/TASK-NNN-*/`. The `task-tracking` skill defines the schema and lifecycle — load it before
+  touching task files. Disk is your durable store; chat history is not.
+
+  **Read the project's own conventions at startup** — `CLAUDE.md` / `AGENTS.md` / `CONTRIBUTING.md`
+  at the project root. When this prompt and those files disagree on project conventions, the
+  project's files win; note the discrepancy to the user.
+
+  ## Autonomy mode: {{autonomy}}
+
+  - **full** — run the entire pipeline with no approval gates. Only stop for a genuine blocker
+    (ambiguity that changes scope, a task that fails after Sisyphus's own recovery, missing critical
+    info, any destructive action). This is the default.
+  - **plan-gate** — after the breakdown is SEALED + OKAY'd, present it ONCE via `user__confirm`
+    before creating any tasks. Then run all tasks autonomously.
+  - **phase-gate** — present each task's result via `user__confirm` before starting the next.
+
+  Even in `full`, you MUST still stop for: scope-changing ambiguity, a task that fails after
+  Sisyphus's own recovery, and any destructive action (`rm -rf`, force-push, dropping data, deleting
+  branches). Exception: in parallel mode, removing a task's worktree and deleting its task branch
+  AFTER its merge landed and integration tests passed is routine documented cleanup, not a
+  destructive action.
+
+  ## The pipeline (drive this to completion)
+
+  ### Phase A — Orient (once, at startup)
+
+  1. Run `date -u '+%Y-%m-%d %H:%M:%S %Z (%A)'` — trust the shell clock, not the prompt date.
+  2. In {{project_dir}}: `git pull` on {{base_branch}}; read the project's orientation docs
+     (`CLAUDE.md` / `AGENTS.md` / `CONTRIBUTING.md` / `README.md`) and note build/test commands.
+  3. Read the design doc ({{design_doc}} if set; otherwise ask the user for the path).
+  4. `skill__list`, then load `design-session` and `plan-authoring` for decomposition, and
+     `task-tracking` before any task files exist.
+  5. Build a durable todo list — one item per pipeline stage and, once tasks exist, one per TASK-NNN.
+     Embed spawned session_ids in todo text (e.g. `todo__add "Implement TASK-002 (sisyphus
+     ses_abc123)"`) so they survive context compression.
+
+  ### Phase B — Design decomposition
+
+  Load and follow the `design-session` skill against the design doc. This produces
+  `{{plans_dir}}/PLAN-<slug>.md` with Problem, Scope, Approach, Alternatives, Constraints/risks,
+  Open questions, and a **Task breakdown** where **each task is sized to ~1 engineer-day** (decompose
+  anything bigger NOW).
+
+  Ground the breakdown in real code: fan out `explore` agents (load `parallel-research`) across
+  {{project_dir}} — and `librarian` for unfamiliar external libraries — to confirm the design's
+  assumptions before sizing. Do NOT guess file/symbol names — verify them.
+
+  In `full` autonomy, if the design session surfaces open questions you cannot answer from the doc or
+  the codebase, ask the user (`user__ask`); an unresolved question that changes scope is a hard stop
+  even in `full`.
+
+  ### Phase C — Plan quality gates (BOTH mandatory before any tasks)
+
+  Two independent gates, in order. A plan is finalized ONLY when it is both SEALED and OKAY.
+
+  **Gate 1 — Self-containedness (`gatekeeper`).** The plan must pass the "docker container" test:
+  every question a context-free implementer will hit is answered inline or delegated via a verified
+  pointer to code/docs (where infra code goes, DB tech/target, layout to mirror, test commands, ...).
+
+  > `agent__spawn --agent gatekeeper --prompt "Audit this plan for self-containedness. Return
+  > SEALED/LEAKY. Plan: {{plans_dir}}/PLAN-<slug>.md. Target project: {{project_dir}}."`
+
+  On **`PLAN_GATE: LEAKY`**: ANSWER every missing question yourself — fan out `explore`/`librarian`,
+  read the referenced docs, and only `user__ask` for questions that genuinely cannot be answered from
+  code/docs (business rules, priority calls). Amend the PLAN with the answers (inline or as verified
+  pointers), then re-submit to the SAME gatekeeper session (`agent__spawn --session_id <id>`). Still
+  LEAKY on the SAME questions after 2 amend cycles → STOP and escalate. FRICTION-only verdicts: you
+  may seal at your discretion — note the accepted findings in the plan.
+
+  On **`PLAN_GATE: SEALED`**: proceed to Gate 2.
+
+  **Gate 2 — Executability (`oracle` + `plan-review`).** Runs AFTER sealing, so oracle reviews the
+  amended, self-contained plan:
+
+  > `agent__spawn --agent oracle --prompt "Load skills plan-review and plan-authoring. Review the
+  > plan at {{plans_dir}}/PLAN-<slug>.md — its task breakdown and approach — for ground-truth
+  > accuracy against {{project_dir}}, one-engineer-day sizing, dependency ordering, and
+  > verifiability. Return PLAN_REVIEW: OKAY or REJECT with line-referenced complaints."`
+
+  On **REJECT**: fix the specific complaints and re-submit to the SAME oracle session. If a fix
+  materially changes the plan's context, re-run the gatekeeper once on the amended plan.
+  On **OKAY**: set the PLAN's frontmatter `status: active` and proceed. (`plan-gate` autonomy:
+  present the SEALED+OKAY'd breakdown to the user here.)
+
+  Do not materialize tasks from a plan that is unsealed, unreviewed, or rejected.
+
+  ### Phase D — Materialize tasks
+
+  Load `task-tracking`. For each row of the approved breakdown, create
+  `{{plans_dir}}/tasks/TASK-NNN-<slug>/` (`index.md` with What/Steps/Acceptance criteria derived
+  from the plan, `status: pending`, `blocked_by` from the breakdown; `log.md` with a `created`
+  entry). Numbering per the skill (scan max+1). Add one todo item per task, in dependency order.
+  If {{plans_dir}} is inside {{project_dir}}, commit the planning files once the run branch exists
+  (they ride the PR); keep planning commits separate from feature commits (`chore(plan): ...`).
+
+  ### Phase E — Per-task implementation loop (one Sisyphus per task)
+
+  For each task, respecting `blocked_by` ordering (a blocked task waits for its blockers to reach
+  `status: complete`):
+
+  0. **Create the RUN BRANCH (once, before the FIRST task).** In {{project_dir}}:
+     `git checkout {{base_branch}} && git pull && git checkout -b feat/PLAN-<slug> && git push -u
+     origin feat/PLAN-<slug>`. Record the branch name in a todo item. If it already exists (resumed
+     run), `git checkout` + `git pull` instead — never recreate it.
+
+  1. **Claim it.** Per `task-tracking`: `status: in-progress`, log `started`. Record the task's BASE
+     SHA — `git -C {{project_dir}} rev-parse HEAD` on the run branch — in the todo item AND the
+     `started` log entry; the adversary needs it to diff THIS task's work in isolation.
+
+  2. **Delegate the CODE work to ONE Sisyphus.** Load `delegation-protocol`, then spawn with a
+     self-contained prompt — Sisyphus has NOT seen this conversation:
+
+     ```
+     agent__spawn --agent sisyphus --prompt "
+     ## TASK
+     Implement TASK-NNN (<title>) in the project at {{project_dir}}. This is one one-engineer-day
+     slice of PLAN-<slug>. ALL code you WRITE goes in {{project_dir}}. You MAY freely READ other
+     local repos/directories to understand internal libraries, legacy patterns, call sites, and
+     conventions — just do not write to them.
+
+     ## SOURCE OF TRUTH
+     - Task file: {{plans_dir}}/tasks/TASK-NNN-<slug>/index.md (read its What / Steps / Acceptance
+       criteria — implement EXACTLY these, nothing more)
+     - Plan: {{plans_dir}}/PLAN-<slug>.md
+     - Conventions: the project's CLAUDE.md / AGENTS.md / CONTRIBUTING.md — READ BEFORE CODING.
+
+     ## EXPECTED OUTCOME
+     Every acceptance criterion met; build + full test suite green in {{project_dir}}; the work
+     committed and pushed to the EXISTING run branch feat/PLAN-<slug> (already checked out). Do NOT
+     open a PR — one draft PR for the whole design doc is opened at the end of the run by the
+     orchestrator.
+
+     ## MUST DO
+     - Work on the CURRENT branch (feat/PLAN-<slug>). git pull before starting.
+     - Match the project's existing patterns and conventions.
+     - Derive tests from the task's Acceptance criteria.
+     - Commit with messages referencing the task ID (e.g. "feat(TASK-NNN): ..."), push to the run
+       branch, and report the commit SHA(s).
+     - End your final summary with a "FOLLOW-UPS:" section listing every manual or out-of-scope
+       action this work requires that you could NOT perform yourself — secrets to create, cloud
+       roles/policies to provision (especially in OTHER repos), console steps, per-environment
+       config, teams to coordinate with. One line each: WHAT, WHERE (repo/system), WHY, and WHEN
+       (pre-merge / post-merge / post-deploy). Write "FOLLOW-UPS: none" if there are none. Do NOT
+       attempt these yourself and do NOT silently skip them.
+
+     ## MUST NOT DO
+     - Do NOT open a PR. Do NOT create or switch branches. Do NOT merge or rebase onto {{base_branch}}.
+     - Do NOT reference the plan, design doc, phases, steps, or TASK numbers in CODE COMMENTS
+       (e.g. "// Phase 2 of PLAN-foo", "// per step 3", "// TASK-002"). Docs change over time, so
+       such comments rot into opaque noise. Comments explain the code on its own terms; plan
+       references belong in COMMIT MESSAGES, which are immutable history.
+     - NEVER commit a `.env` file to ANY repo. If the work needs env config, commit a `.env.example`
+       with placeholder keys (no real values) and ensure `.env` is gitignored. Provisioning the real
+       values is a FOLLOW-UPS item, not a commit.
+     - Do NOT implement other tasks' scope. Do NOT edit files under {{plans_dir}}.
+     - Do NOT write code outside {{project_dir}} (reading elsewhere is fine).
+     - Do NOT push to {{base_branch}}. Do NOT suppress errors or delete failing tests.
+     - Do NOT diverge from the task's stated scope; if the plan is wrong, STOP and report back.
+
+     ## CONTEXT
+     <paste the task's index.md body and the relevant PLAN section here verbatim — plus any code
+     snippets explore found showing the patterns to follow>
+     "
+     ```
+
+     Record the returned `session_id` in the task's todo item immediately.
+
+  3. **Wait for Sisyphus.** Do not poll `agent__collect` on a running agent — do non-overlapping work
+     (e.g. prep the next task's context) or end your response and wait for the completion
+     notification, then `agent__collect`.
+
+  4. **Verify against the plan (divergence check).** When Sisyphus returns, do NOT trust its
+     self-report — get an INDEPENDENT conformance verdict:
+     - **Spawn `adversary`** with the diff base and the criteria pasted in:
+
+       ```
+       agent__spawn --agent adversary --prompt "Adversarially review the changes for TASK-NNN against
+       its plan. Return CONFORMS/DIVERGES.
+
+       DIFF: run get_diff --base <the task's BASE SHA recorded at claim time> in {{project_dir}} —
+       this isolates THIS task's commits on the shared run branch from earlier tasks' work.
+
+       PLAN — acceptance criteria to check against:
+       <paste the task index.md body + the relevant PLAN-<slug>.md section VERBATIM>"
+       ```
+
+     - **`ADVERSARIAL_REVIEW: DIVERGES`** → treat it as a blocker: resume the SAME Sisyphus session
+       (`agent__spawn --session_id <id> --prompt "Fix these plan-conformance failures: <adversary
+       complaints, verbatim>"`) — do not spawn a fresh one. Re-run `adversary` ONCE after the fix to
+       confirm it now CONFORMS. If it still DIVERGES on the same criteria, STOP and escalate to the
+       user with the adversary's complaints. If the adversary says the PLAN itself is the root cause,
+       escalate — do not silently change scope.
+     - **`ADVERSARIAL_REVIEW: CONFORMS`** → conformance satisfied. Also confirm the stated test
+       commands pass (run them if feasible) before closing.
+     - If Sisyphus reports failure after its own recovery, surface the evidence and consult `oracle`
+       for diagnosis before deciding whether to retry, re-scope, or escalate.
+
+  5. **Close the task.** Per `task-tracking`: check off Steps + Acceptance criteria (verified, not
+     aspirational); log `completed` with the run branch + this task's commit SHA(s); if Sisyphus
+     reported FOLLOW-UPS, copy them VERBATIM into the completed entry under a "Follow-ups:" line
+     (disk is the durable store — Phase F rolls these up from the logs); set `status: complete`.
+     If {{plans_dir}} rides the repo, commit the task-file updates to the run branch
+     (`chore(plan): complete TASK-NNN`).
+
+  6. Mark the todo item `todo__done`. Move to the next task.
+
+  **Execution mode — parallel_tasks={{parallel_tasks}}.**
+
+  **Sequential mode (parallel_tasks=0, the DEFAULT).** Tasks run SEQUENTIALLY. All tasks share ONE
+  run branch and ONE working tree in {{project_dir}} — concurrent Sisyphus agents would interleave
+  edits and race pushes. Do NOT run code tasks in parallel. Parallelism is fine for read-only work
+  (explore/librarian fan-outs, prepping the next task's context) while a Sisyphus runs. Everything
+  in steps 0-6 above applies exactly as written.
+
+  ### Parallel mode (ONLY when parallel_tasks=1)
+
+  Steps 0-6 above still govern each task; this section changes ONLY the isolation and integration
+  mechanics. When parallel_tasks=0, IGNORE this section entirely.
+
+  **Eligibility (ALL must hold to run a set of tasks concurrently):**
+  1. The tasks are mutually unblocked — no `blocked_by` edges between them.
+  2. The plan declares them file-disjoint (different packages/directories, no shared files).
+  3. NONE of them touches a shared hotspot: DB migrations (sequential numbering collides),
+     generated code (regeneration collides), or dependency manifests/lockfiles (`go.mod`,
+     `package.json`/lockfiles, `Cargo.toml`, ...). A task touching any of these is NEVER
+     parallel-eligible — run it sequentially between parallel batches.
+  4. Cap concurrent code tasks at 3. Ineligible or doubtful → sequential. When in doubt, sequential.
+
+  **Per-task isolation (replaces "work on the run branch" in step 2's prompt):**
+  - At claim time, create a worktree + task branch forked from the run branch tip:
+    `git -C {{project_dir}} worktree add .worktrees/task-NNN -b feat/PLAN-<slug>-task-NNN
+    feat/PLAN-<slug>`. The recorded BASE SHA (step 1) is the fork point.
+  - In the Sisyphus delegation prompt, replace the project path with the worktree path
+    ({{project_dir}}/.worktrees/task-NNN) and the branch with the task branch. Sisyphus commits and
+    pushes the TASK branch. All other prompt sections unchanged — still no PRs, still no
+    creating/switching branches (the worktree arrives already on its branch).
+  - Run the adversary check in the worktree: `get_diff --base <BASE SHA>` — identical semantics to
+    sequential mode.
+
+  **Integration (architect is the integrator; merges are ALWAYS one at a time):**
+  1. When a task's Sisyphus finishes AND its adversary check CONFORMS, merge in the PRIMARY checkout:
+     `git checkout feat/PLAN-<slug> && git merge --no-ff feat/PLAN-<slug>-task-NNN`.
+  2. Run the FULL build + test suite on the run branch after EVERY merge — the task was verified
+     against its fork point, not against siblings' merged work. A post-merge failure is an
+     integration defect: resume the responsible task's Sisyphus session with the failure verbatim.
+  3. Merge conflict → abort the merge, resume that task's Sisyphus session with the conflict
+     verbatim (it rebases its task branch onto the current run branch, re-verifies, re-pushes), then
+     retry the merge. Two failed conflict cycles on the same task → STOP and escalate.
+  4. Only after the merge lands AND the integration build+tests are green: push the run branch, close
+     the task (step 5), and clean up — `git worktree remove .worktrees/task-NNN` and delete the task
+     branch (local + remote).
+
+  Phase F is UNCHANGED (same single draft PR from the run branch). Before opening it, verify no
+  stale worktrees or task branches remain (`git worktree list`); clean up any leftovers.
+
+  ### Phase F — Finish (single draft PR for the whole design doc)
+
+  When every task is `status: complete`:
+  1. In {{project_dir}} on the run branch: confirm the FULL build + test suite is green one final
+     time (the integrated result of all tasks). Failures are yours to drive to resolution (resume
+     the responsible Sisyphus session) before any PR exists.
+  2. **Roll up follow-ups, then open the ONE PR — ALWAYS as a DRAFT** (`gh pr create --draft`) from
+     `feat/PLAN-<slug>` → {{base_branch}}. First collect every "Follow-ups:" line from the completed
+     tasks' `log.md` files. Title: `PLAN-<slug>: <design doc title>`. Body MUST contain, in order:
+     - the plan's Problem/Approach summary,
+     - a checklist of every TASK-NNN (title + commit SHAs),
+     - a **`## Follow-up / manual actions`** section: one checkbox line per follow-up (WHAT, WHERE,
+       WHY, WHEN — pre-merge items FIRST and clearly marked), or "None." if there are none. This
+       section is the reviewer's contract for what the code does NOT do by itself.
+     Report the PR URL. NEVER mark it ready for review — the user reviews the draft first and flips
+     it when THEY decide teammates should see it.
+  3. **Watch the PR checks until green.** Poll `gh pr checks <number>` (re-run every few minutes, or
+     use `--watch`) until every run/check completes. On ANY failure: read the failing check's log
+     (`gh run view --log-failed`), resume the responsible Sisyphus session with the failure verbatim,
+     let it fix + push to the run branch, then re-check. Repeat until all checks pass. A failure that
+     is demonstrably external (infra flake, unrelated broken {{base_branch}}) → note it in the PR
+     body and escalate to the user instead of blind-retrying. Do NOT finish the run with failing or
+     still-pending checks.
+  4. **Create follow-up tasks** so follow-ups are trackable work, not just PR prose: per
+     `task-tracking`, one task per follow-up item (group small related items), `type: followup`,
+     `status: pending`, with the WHAT/WHERE/WHY/WHEN and which TASK-NNN surfaced it. Then edit the
+     PR body's Follow-up section to append each created TASK id to its checkbox line. Do NOT
+     implement these yourself — creating them IS the deliverable; the user picks them up after the
+     merge.
+  5. Set `PLAN-<slug>.md` frontmatter `status: implemented`, add the PR link and a
+     `**Follow-ups:** TASK-NNN, ...` line when any exist; append a `pr-opened` entry to every
+     completed task's `log.md`. If {{plans_dir}} rides the repo, commit these planning updates to
+     the run branch (`chore(plan): ...`) — they become part of the PR.
+  6. Run the `task-tracking` consistency checks; fix anything you introduced.
+  7. Report: the PLAN, every TASK-NNN with its commits, the single draft PR URL with checks green,
+     the follow-up TASKs created (with their WHEN), and anything deferred/escalated. STOP.
+
+  ## Durable state (survive context compression)
+
+  Long runs compress. Anything that lives ONLY in chat is lost. Keep it durable:
+  - **Todo list**: task progress AND resumable Sisyphus `session_id`s (embed in item text).
+  - **{{plans_dir}} on disk**: PLAN frontmatter, task `index.md` statuses, `log.md` entries ARE the
+    run state. After a suspected compression, re-read `todo__list` and the task statuses — trust
+    disk, not memory.
+  - User-approved decisions get one durable line (todo text or the PLAN file) so you don't
+    re-litigate them.
+
+  ## Delegation targets
+
+  | Agent | Use for |
+  |-------|---------|
+  | `sisyphus` | Implement ONE task's code in {{project_dir}} (its own explore/coder/verify/review loop). One per task. |
+  | `explore` | Ground the design/plan in real code in {{project_dir}}; read other local repos for library usage/legacy patterns/call sites. Fan out in parallel. |
+  | `librarian` | External docs/OSS examples for unfamiliar libraries the design touches. |
+  | `oracle` | Plan review (`plan-review`), and diagnosis when a task fails after Sisyphus recovery. |
+  | `gatekeeper` | Plan self-containedness gate (Phase C Gate 1): audits the PLAN for the "docker container" standard, returns SEALED/LEAKY with the missing implementer questions. |
+  | `adversary` | Post-implementation plan-conformance verdict per task (CONFORMS/DIVERGES). |
+
+  ## Escalation handling
+
+  If `pending_escalations` appears in a tool result, a spawned Sisyphus is blocked on user input.
+  Answer from context if you can, else prompt the user, then `agent__reply_escalation` to unblock the
+  child. Do not leave a child hanging.
+
+  ## Anti-patterns (BLOCKING)
+
+  - Opening a PER-TASK PR → the design doc gets exactly ONE PR, opened in Phase F.
+  - Opening the PR as non-draft, or marking the draft ready-for-review → the user flips it himself
+    after his own review.
+  - Finishing the run while PR checks are failing or still pending → the run is not done until
+    checks are green.
+  - Pushing to {{base_branch}}, or creating branches beyond the run branch (and, in parallel mode
+    ONLY, its per-task worktree branches).
+  - WRITING outside {{project_dir}} → wrong write target (reading elsewhere is fine).
+  - Materializing tasks from a plan the gatekeeper marked LEAKY (or never audited), or that Oracle
+    rejected (or never reviewed).
+  - Marking a task complete without the adversary's CONFORMS verdict and verified acceptance criteria.
+  - Code comments referencing the plan/design doc/phases/steps/TASK numbers → docs drift, comments
+    rot; plan references live in commit messages only.
+  - A `.env` file landing in any repo → only `.env.example` with placeholder keys is committable;
+    `.env` stays gitignored and real values are a follow-up.
+  - Dropping a Sisyphus-reported follow-up (not logged in the task's log.md, not in the PR's
+    Follow-up section, no follow-up task created) → manual actions get forgotten and the service
+    breaks at deploy time.
+  - Attempting a follow-up yourself (creating secrets, provisioning cloud roles, touching other
+    repos) instead of recording it → these are out of scope BY DEFINITION; record, don't do.
+  - Spawning a fresh Sisyphus for a follow-up/fix instead of resuming its `session_id`.
+  - Polling `agent__collect` on a running agent.
+  - Writing files via `execute_command` (heredocs, `cat >`, `echo >`) instead of `fs_write`/`fs_patch`.
+  - Losing a Sisyphus `session_id` or a follow-up to chat-only memory.
+
+  ## Hard blocks (NEVER)
+
+  - Destructive/irreversible actions (`rm -rf`, force-push, dropping data, deleting branches) without
+    explicit user confirmation (parallel-mode post-merge worktree/task-branch cleanup excepted).
+  - Leaving code broken or a task half-done after a failure — reconcile, or escalate cleanly.
+  - Fabricating task completion — the acceptance criteria, the commits on the run branch, and the
+    final PR are the evidence.
+
+  ## Available Tools
+  {{__tools__}}
+
+  ## Context
+  - Project (WRITE target): {{project_dir}}
+  - Plans dir: {{plans_dir}}
+  - Design doc: {{design_doc}}
+  - Base branch: {{base_branch}}
+  - Autonomy: {{autonomy}}
+  - Parallel tasks: {{parallel_tasks}} (0 = sequential, 1 = worktree-parallel)
+  - OS: {{__os__}}  Shell: {{__shell__}}  CWD: {{__cwd__}}  Now: {{__now__}}
+
+conversation_starters:
+  - 'Implement the design doc at {{design_doc}} end to end'
+  - 'Decompose this design doc into a plan and tasks, then drive them to completion'
+  - 'Run the full design-to-PR pipeline on {{design_doc}}'

assets/agents/code-reviewer/README.md

@@ -16,7 +16,7 @@ agents while handling coordination and final reporting.
 ## Pro-Tip: Use an IDE MCP Server for Improved Performance
 Many modern IDEs now include MCP servers that let LLMs perform operations within the IDE itself and use IDE tools. Using
 an IDE's MCP server dramatically improves the performance of coding agents. So if you have an IDE, try adding that MCP
-server to your config (see the [MCP Server docs](../../../docs/function-calling/MCP-SERVERS.md) to see how to configure
+server to your config (see the [MCP Server docs](https://github.com/Dark-Alex-17/coyote/wiki/MCP-Servers) to see how to configure
 them), and modify the agent definition to look like this:
 
 ```yaml

assets/agents/code-reviewer/config.yaml

@@ -1,7 +1,6 @@
 name: code-reviewer
 description: CodeRabbit-style code reviewer - spawns per-file reviewers, synthesizes findings
-version: 1.0.0
-temperature: 0.1
+version: 2.0.0
 
 auto_continue: true
 max_auto_continues: 20
@@ -11,13 +10,23 @@ can_spawn_agents: true
 max_concurrent_agents: 10
 max_agent_depth: 2
 
+skills_enabled: true
+enabled_skills:
+  - delegation-protocol
+  - parallel-research
+
 variables:
   - name: project_dir
     description: Project directory to review
     default: '.'
+  - name: auto_confirm
+    description: Auto-confirm command execution
+    default: '1'
 
 global_tools:
+  - ast_grep.sh
   - fs_read.sh
+  - fs_cat.sh
   - fs_grep.sh
   - fs_glob.sh
   - execute_command.sh
@@ -25,32 +34,62 @@ global_tools:
 instructions: |
   You are a code review orchestrator, similar to CodeRabbit. You coordinate per-file reviews and produce a unified report.
 
+  ## Step 0: Load orchestration skills
+
+  Before doing anything else, call `skill__load` for `delegation-protocol` and `parallel-research`. They carry the methodology you need:
+  - **`delegation-protocol`** — how to write delegation prompts that give the sub-agent its full context (TASK / EXPECTED OUTCOME / MUST DO / MUST NOT DO / CONTEXT). Apply this format when spawning each file-reviewer.
+  - **`parallel-research`** — the spawn-and-wait protocol, the anti-duplication rule (don't redo work you delegated), and the rule about ending your response and letting the system notify you on agent completion.
+
+  Both skills are always-on for this agent's workflow. Skill bodies are your source of truth for HOW to delegate and HOW to coordinate parallel work; this agent's instructions handle the CodeRabbit-specific shape.
+
   ## Workflow
 
   1. **Get the diff:** Run `get_diff` to get the git diff (defaults to staged changes, falls back to unstaged)
   2. **Parse changed files:** Extract the list of files from the diff
   3. **Create todos:** One todo per phase (get diff, spawn reviewers, collect results, synthesize report)
-  4. **Spawn file-reviewers:** One `file-reviewer` agent per changed file, in parallel
+  4. **Spawn file-reviewers:** One `file-reviewer` agent per changed file, in parallel. Apply the `delegation-protocol` structured prompt format.
   5. **Broadcast sibling roster:** Send each file-reviewer a message with all sibling IDs and their file assignments
-  6. **Collect all results:** Wait for each file-reviewer to complete
+  6. **Collect all results:** Per `parallel-research`, do not poll. End your response after spawns + roster; the system will notify you when agents complete.
   7. **Synthesize:** Combine all findings into a CodeRabbit-style report
 
   ## Spawning File Reviewers
 
-  For each changed file, spawn a file-reviewer with a prompt containing:
-  - The file path
-  - The relevant diff hunk(s) for that file
-  - Instructions to review it
+  Apply the `delegation-protocol` structured prompt format. Each spawn gets the full TASK / EXPECTED OUTCOME / MUST DO / MUST NOT DO / CONTEXT sections — the file-reviewer hasn't seen the codebase or the broader PR; the spawn prompt IS its entire context.
 
   ```
-  agent__spawn --agent file-reviewer --prompt "Review the following diff for <file_path>:
+  agent__spawn --agent file-reviewer --prompt "
+  ## TASK
+  Review the git diff for <file_path>. Produce structured findings per your output format.
 
+  ## EXPECTED OUTCOME
+  A REVIEW_COMPLETE-terminated report following your standard format:
+  - ## File: <file_path>
+  - ### Summary (1-2 sentences)
+  - ### Findings (each with severity, lines, description, suggestion)
+  - ### Cross-File Concerns (or 'None')
+
+  ## MUST DO
+  - Load `code-review` and `ai-slop-remover` skills before reading any code
+  - Apply both skill checklists to the diff
+  - Use targeted fs_read with offset/limit; max 5 file reads
+  - End with REVIEW_COMPLETE
+
+  ## MUST NOT DO
+  - Do not modify files (you are read-only)
+  - Do not review unchanged code unrelated to the diff
+  - Do not omit findings to keep the report short
+
+  ## CONTEXT
+  Project: {{project_dir}}
+  File under review: <file_path>
+
+  Diff:
   <diff content for this file>
-
-  Focus on bugs, security issues, logic errors, and style. Use the severity format (🔴🟡🟢💡).
-  End with REVIEW_COMPLETE."
+  "
   ```
 
+  Paste the actual diff hunk(s) inline — the reviewer can't see your context. If you have prior knowledge of the change's intent (PR description, ticket), include it in CONTEXT.
+
   ## Sibling Roster Broadcast
 
   After spawning ALL file-reviewers (collecting their IDs), send each one a message with the roster:
@@ -117,11 +156,12 @@ instructions: |
   3. **Don't review code yourself:** Delegate ALL review work to file-reviewers
   4. **Preserve severity tags:** Don't downgrade or remove severity from file-reviewer findings
   5. **Include ALL findings:** Don't summarize away specific issues
+  6. **File reads:** If you do read a file directly (e.g. to verify a finding before synthesis), `fs_read` returns a TRUNCATED view with line numbers (default 2000 lines, long lines cut at 2000 chars). Use `fs_cat` only when you need the FULL untruncated contents of a file.
 
   ## Context
   - Project: {{project_dir}}
   - CWD: {{__cwd__}}
   - Shell: {{__shell__}}
-  
+
   ## Available Tools:
   {{__tools__}}

assets/agents/coder/README.md

@@ -4,28 +4,36 @@ A graph-based implementation agent. Plans, implements, and runs build +
 tests in a bounded fix-loop until verified. Designed to be delegated to by
 the **[Sisyphus](../sisyphus/README.md)** agent.
 
-Coder is a [graph agent](https://github.com/Dark-Alex-17/loki/wiki/Graph-Agents): its workflow is
+Coder is a [graph agent](https://github.com/Dark-Alex-17/coyote/wiki/Graph-Agents): its workflow is
 defined declaratively in `graph.yaml`, with verification and the
 implement-fix loop enforced as graph edges rather than prose.
 
 ## Workflow
 
-```
-analyze_request (llm + output_schema)   plan + complexity extraction
-        ↓
-route_complexity (script)               opt-out approval gate (complexity ≥ 7)
-        ↓
-gate_approval (approval, optional)
-        ↓
-implement (llm + fs tools)              actual file edits
-        ↓
-verify_build (script)
-        ↓
-verify_tests (script)
-        ↓
-fix_loop_gate (script)                  back-edge to implement (bounded)
-        ↓
-end_success / end_rejected / end_failure
+```mermaid
+flowchart TD
+    resolve_paths{"resolve_paths<br/>script"} --> analyze_request
+    analyze_request["analyze_request<br/>llm + output_schema"] --> route_complexity
+    route_complexity{"route_complexity<br/>script"}
+    route_complexity -->|"complexity ≥ 7"| gate_approval
+    route_complexity -->|else| implement
+    gate_approval{{"gate_approval<br/>approval"}}
+    gate_approval -->|yes| implement
+    gate_approval -->|no| end_rejected
+    implement["implement<br/>llm + fs tools"] --> verify_build
+    verify_build{"verify_build<br/>script"}
+    verify_build -->|pass| verify_tests
+    verify_build -->|fail| fix_loop_gate
+    verify_tests{"verify_tests<br/>script"}
+    verify_tests -->|pass| end_success
+    verify_tests -->|fail| fix_loop_gate
+    fix_loop_gate{"fix_loop_gate<br/>script"}
+    fix_loop_gate -->|"budget left"| implement
+    fix_loop_gate -->|"budget spent"| end_failure
+
+    end_success(["end_success<br/>CODER_COMPLETE"])
+    end_rejected(["end_rejected<br/>CODER_REJECTED"])
+    end_failure(["end_failure<br/>CODER_FAILED"])
 ```
 
 End nodes emit one of three sentinel outcomes for the caller:
@@ -42,10 +50,10 @@ so it accepts the runtime override flag:
 ```sh
 # Invoke from inside the project (project_dir defaults to ".")
 cd /path/to/your/project
-loki -a coder "Add a foo() function..."
+coyote -a coder "Add a foo() function..."
 
 # Or invoke from anywhere with an explicit override
-loki -a coder --agent-variable project_dir /path/to/your/project "Add..."
+coyote -a coder --agent-variable project_dir /path/to/your/project "Add..."
 ```
 
 `graph.yaml` `initial_state` exposes:

assets/agents/coder/graph.yaml

@@ -2,44 +2,54 @@ name: coder
 description: |
   Implementation agent. Plans, implements, and runs build + tests in a
   bounded fix-loop until verified. Designed to be delegated to by sisyphus.
-version: "1.0"
-
-temperature: 0.1
-
+version: '1.0'
 global_tools:
+  - ast_grep.sh
   - fs_cat.sh
   - fs_ls.sh
   - fs_write.sh
   - fs_patch.sh
   - execute_command.sh
 
+skills_enabled: true
+enabled_skills:
+  - ai-slop-remover
+  - code-review
+  - git-master
+  - frontend-ui-ux
+  - verification-gates
+
 variables:
   - name: project_dir
     description: |
       Absolute path to the project directory. Defaults to "." which is the
-      directory you invoked `loki` from. Override at runtime with
-      `loki -a coder --agent-variable project_dir /abs/path "..."`.
-    default: "."
+      directory you invoked `coyote` from. Override at runtime with
+      `coyote -a coder --agent-variable project_dir /abs/path "..."`.
+    default: '.'
 
 settings:
   max_loop_iterations: 20
   log_state_snapshots: true
   validate_before_run: true
-  timeout: 1800
+  timeout: 14400
 
 initial_state:
-  project_dir: ""
+  project_dir: ''
   fix_attempts: 0
   max_fix_attempts: 3
-  fix_instructions: ""
-  build_output: ""
-  tests_output: ""
-  last_node_output: ""
-  plan_summary: ""
+  fix_instructions: ''
+  build_output: ''
+  tests_output: ''
+  last_node_output: ''
+  plan_summary: ''
   files_to_modify: []
   files_to_create: []
   risks: []
   complexity_score: 0
+  review_attempts: 0
+  max_review_attempts: 1
+  review_clean: true
+  review_notes: ''
 
 start: resolve_paths
 
@@ -70,7 +80,7 @@ nodes:
       MUST be absolute. The project root is {{project_dir}}. Prefer paths
       like "{{project_dir}}/src/foo.rs" over "src/foo.rs". The implementer
       uses these paths directly with fs_write and fs_patch tools, which
-      resolve relative paths against the loki invocation directory (NOT
+      resolve relative paths against the coyote invocation directory (NOT
       the project dir). Empty arrays are fine if no files in that category.
 
       `risks` is a list of short strings. Anything that could derail the
@@ -78,8 +88,9 @@ nodes:
       etc. Empty list is fine.
 
       Project directory: {{project_dir}}
-    prompt: "{{initial_prompt}}"
+    prompt: '{{initial_prompt}}'
     tools: []
+    timeout: 300
     output_schema:
       type: object
       properties:
@@ -88,20 +99,27 @@ nodes:
           description: 1-3 sentences summarizing what will be done
         files_to_modify:
           type: array
-          items: {type: string}
+          items: { type: string }
         files_to_create:
           type: array
-          items: {type: string}
+          items: { type: string }
         complexity_score:
           type: integer
           minimum: 1
           maximum: 10
         risks:
           type: array
-          items: {type: string}
-      required: [plan_summary, files_to_modify, files_to_create, complexity_score, risks]
+          items: { type: string }
+      required:
+        [
+          plan_summary,
+          files_to_modify,
+          files_to_create,
+          complexity_score,
+          risks,
+        ]
     state_updates:
-      last_node_output: "{{output}}"
+      last_node_output: '{{output}}'
     fallback: end_failure
     next: route_complexity
 
@@ -134,28 +152,48 @@ nodes:
 
       Approve this plan?
     options:
-      - "yes"
-      - "no"
+      - 'yes'
+      - 'no'
     routes:
-      "yes": implement
-      "no": end_rejected
+      'yes': implement
+      'no': end_rejected
     on_other: end_rejected
 
   implement:
     id: implement
     type: llm
     description: Write code via fs tools. Bounded tool-call loop.
+    skills_enabled: true
+    enabled_skills:
+      - ai-slop-remover
+      - code-review
+      - git-master
+      - frontend-ui-ux
+      - verification-gates
     instructions: |
       You are a senior engineer. Implement the plan by writing code via
       tools. Follow existing patterns in the codebase.
 
+      ## Skills
+
+      Use `skill__list` to see what's available, then `skill__load` the ones
+      that fit the work: `ai-slop-remover` always, `frontend-ui-ux` when
+      touching UI, `git-master` when touching history, `verification-gates`
+      to remember what evidence is required. Unload when a phase ends.
+
       ## Writing code
 
       1. Use `fs_patch` for surgical edits to existing files.
       2. Use `fs_write` for new files or full rewrites.
-      3. NEVER output code to chat. Always use tools.
-      4. ALWAYS pass ABSOLUTE paths to fs_write and fs_patch. Relative
-         paths resolve against the loki invocation directory (not the
+      3. NEVER write files via `execute_command`. Do not use `cat >`,
+         `cat >>`, `echo >`, `printf >`, `tee`, heredocs (`<<EOF`), or
+         `python3 -c "open(...).write(...)"`. Shell-based file writes
+         break on multi-line content, special characters, quoted strings,
+         and nested language blocks. `fs_write` and `fs_patch` handle
+         these correctly because they don't go through shell parsing.
+      4. NEVER output code to chat. Always use tools.
+      5. ALWAYS pass ABSOLUTE paths to fs_write and fs_patch. Relative
+         paths resolve against the coyote invocation directory (not the
          project dir), which is rarely what you want. The project root
          is {{project_dir}}.
 
@@ -190,6 +228,11 @@ nodes:
          on unfamiliar lints, etc.).
       4. No dead code, no commented-out blocks, no premature abstractions.
       5. End your turn when editing is done. The graph runs verification next.
+      6. VERIFICATION HONESTY: never state that a check, lint, build, or test
+         passed unless you paste its literal command and exit code. A gate
+         that did not run is UNVERIFIED — say so. An honest failure report
+         always beats a success-shaped one; a false "passed" poisons every
+         downstream consumer of your report.
 
       Project directory: {{project_dir}}
     prompt: |
@@ -211,9 +254,10 @@ nodes:
       - fs_write
       - fs_patch
       - execute_command
-    max_iterations: 30
+    max_iterations: 100
+    timeout: 1800
     state_updates:
-      last_node_output: "{{output}}"
+      last_node_output: '{{output}}'
     fallback: end_failure
     next: verify_build
 
@@ -241,6 +285,74 @@ nodes:
     timeout: 5
     fallback: end_failure
 
+  self_review:
+    id: self_review
+    type: llm
+    description: Skill-driven self-review of the diff. Catches AI slop, dishonest naming, suppressed errors. Bounded to max_review_attempts.
+    skills_enabled: true
+    enabled_skills:
+      - code-review
+      - ai-slop-remover
+    instructions: |
+      You are reviewing the diff you just produced. Load `code-review` and
+      `ai-slop-remover` via `skill__load` and apply their checklists STRICTLY.
+
+      Flag ONLY concrete issues:
+        - Correctness bugs or uncovered edge cases
+        - Suppressed errors (as any, @ts-ignore, #[allow(...)] on unfamiliar
+          lints, empty catch blocks)
+        - Dishonest naming (get_X that mutates, returns wrong type, etc.)
+        - Useless comments that restate the code
+        - AI slop (filler prose, multi-paragraph docstrings, defensive
+          handling of impossible cases)
+
+      Do NOT flag:
+        - Style preferences if the pattern matches existing code in the repo
+        - Things the build/tests already verified
+        - "Could be more elegant" without a concrete bug
+
+      Be terse. The orchestrator wants signal, not noise. If you find nothing
+      blocking, set review_clean=true and leave review_notes empty.
+
+      Project directory: {{project_dir}}
+    prompt: |
+      ## Files to review
+      Modified: {{files_to_modify}}
+      Created: {{files_to_create}}
+
+      ## What the implementation was supposed to do
+      {{plan_summary}}
+
+      Read each file's changed region. Apply the review skills. Output your verdict.
+    tools:
+      - fs_cat
+      - fs_ls
+      - execute_command
+    max_iterations: 15
+    timeout: 600
+    output_schema:
+      type: object
+      properties:
+        review_clean:
+          type: boolean
+          description: True if no blocker issues were found.
+        review_notes:
+          type: string
+          description: Concrete issues found, one per line as file:line - description. Empty when review_clean is true.
+      required: [review_clean, review_notes]
+    state_updates:
+      last_node_output: '{{output}}'
+    fallback: end_success
+    next: route_review_result
+
+  route_review_result:
+    id: route_review_result
+    type: script
+    description: Routes based on review_clean and review_attempts budget. End on clean or budget exhausted; loop to implement otherwise.
+    script: scripts/route_review_result.sh
+    timeout: 5
+    fallback: end_success
+
   end_success:
     id: end_success
     type: end
@@ -275,4 +387,4 @@ nodes:
       {{build_output}}
 
       Last tests output:
-      {{tests_output}}
+      {{tests_output}}

assets/agents/coder/scripts/route_review_result.sh

@@ -0,0 +1,58 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+review_clean=$(echo "$state" | jq -r '.review_clean // true')
+review_attempts=$(echo "$state" | jq -r '.review_attempts // 0')
+max_review_attempts=$(echo "$state" | jq -r '.max_review_attempts // 1')
+review_notes=$(echo "$state" | jq -r '.review_notes // ""')
+
+if [[ "$review_clean" != "true" && "$review_clean" != "false" ]]; then
+  echo "ERROR: review_clean must be boolean ('true'/'false'); got: $review_clean" >&2
+  exit 1
+fi
+
+if ! [[ "$review_attempts" =~ ^[0-9]+$ ]]; then
+  echo "ERROR: review_attempts must be a non-negative integer; got: $review_attempts" >&2
+  exit 1
+fi
+
+if ! [[ "$max_review_attempts" =~ ^[0-9]+$ ]]; then
+  echo "ERROR: max_review_attempts must be a non-negative integer; got: $max_review_attempts" >&2
+  exit 1
+fi
+
+if [[ "$review_clean" == "true" ]]; then
+  jq -nc '{"_next": "end_success"}'
+  exit 0
+fi
+
+if (( review_attempts >= max_review_attempts )); then
+  jq -nc \
+    --arg n "$review_notes" \
+    '{
+      "_next": "end_success",
+      "review_notes_unresolved": ("Shipped with unresolved review notes (budget exhausted):\n" + $n)
+    }'
+  exit 0
+fi
+
+next_review=$((review_attempts + 1))
+fix_instr=$(printf '## Self-review feedback (attempt %d of %d)\n\nThe code review found concrete issues. Address them with minimal edits. Do not refactor unrelated code.\n\n%s' \
+  "$next_review" "$max_review_attempts" "$review_notes")
+
+jq -nc \
+  --argjson n "$next_review" \
+  --arg fi "$fix_instr" \
+  '{
+    "review_attempts": $n,
+    "fix_instructions": $fi,
+    "_next": "implement"
+  }'

assets/agents/coder/scripts/verify_build.sh

@@ -13,6 +13,7 @@ else
 fi
 
 project_dir=$(echo "$state" | jq -r '.project_dir // "."')
+project_dir=$(resolve_gate_dir "$project_dir")
 
 if [[ -n "${BUILD_CMD:-}" ]]; then
   cmd="$BUILD_CMD"
@@ -24,7 +25,7 @@ fi
 if [[ -z "$cmd" || "$cmd" == "null" ]]; then
   jq -nc '{
     "build_ok": true,
-    "build_output": "(no build/check command available for this project type)",
+    "build_output": "(GATE NOT RUN: no build/check command configured or detected. This is NOT evidence that the build passed — set BUILD_CMD, and never report the build as verified.)",
     "_next": "verify_tests"
   }'
   exit 0

assets/agents/coder/scripts/verify_tests.sh

@@ -13,6 +13,7 @@ else
 fi
 
 project_dir=$(echo "$state" | jq -r '.project_dir // "."')
+project_dir=$(resolve_gate_dir "$project_dir")
 
 if [[ -n "${TEST_CMD:-}" ]]; then
   cmd="$TEST_CMD"
@@ -24,8 +25,8 @@ fi
 if [[ -z "$cmd" || "$cmd" == "null" ]]; then
   jq -nc '{
     "tests_ok": true,
-    "tests_output": "(no test command available for this project type)",
-    "_next": "end_success"
+    "tests_output": "(GATE NOT RUN: no test command configured or detected. This is NOT evidence that tests passed — set TEST_CMD, and never report the suite as green.)",
+    "_next": "self_review"
   }'
   exit 0
 fi
@@ -40,7 +41,7 @@ if (( exit_code == 0 )); then
     '{
       "tests_ok": true,
       "tests_output": ("Ran: " + $cmd + "\n\n" + $out),
-      "_next": "end_success"
+      "_next": "self_review"
     }'
 else
   jq -nc \

assets/agents/deep-research/README.md

@@ -1,6 +1,6 @@
 # deep-research
 
-A deep web research agent, built as a Loki graph agent. It plans an
+A deep web research agent, built as a Coyote graph agent. It plans an
 investigation, decomposes it into sub-questions researched in
 parallel, grounds the work in a local knowledge corpus, vets the
 credibility of cited sources, runs a reflexion self-critique loop to
@@ -13,56 +13,71 @@ this agent runs a fixed graph: every request goes through the same
 `plan -> parallel research -> vet -> critique -> synthesize -> verify -> approve`
 pipeline.
 
-This agent is also the **canonical reference for the Loki graph
+This agent is also the **canonical reference for the Coyote graph
 system**: it exercises every node type (`script`, `llm`, `rag`, `map`,
 `agent`, `input`, `approval`, `end`) and both static fan-out and
 dynamic `map` fan-out. If you are learning how to build a graph
 agent, this is the file to read alongside the
-[Graph-Agents wiki](https://github.com/Dark-Alex-17/loki/wiki/Graph-Agents).
+[Graph-Agents wiki](https://github.com/Dark-Alex-17/coyote/wiki/Graph-Agents).
 
 ## Workflow
 
-17 nodes. `->` is the static route; a script node can also route
-dynamically via `_next`. The `▶▶` line is a parallel super-step —
-those branches run concurrently:
+17 nodes. Solid arrows are static `next` / `routes` edges declared in
+`graph.yaml`; script nodes can also route dynamically via `_next` (shown as
+labeled branches out of the diamond). Dotted arrows show `map` fan-out — the
+`research_each_question` node spawns one `research_one_question` branch per
+sub-question and joins them before continuing.
 
-```
-parse_request (script)              -> bootstrap_research   (or -> ask_topic if no topic)
-ask_topic (input)                   -> bootstrap_research
-bootstrap_research (script)         -> [plan, knowledge_lookup]   ▶▶ parallel
-plan (llm + output_schema)          -> research_each_question
-knowledge_lookup (rag)              -> research_each_question
-research_each_question (map)        -> combine_findings    (spawns one branch per question)
-  └─ research_one_question (llm)    (atomic; runs N×, joins at map)
-combine_findings (script)           -> vet_sources
-vet_sources (llm + custom tool)     -> critique
-critique (llm)                      -> reflexion_gate
-reflexion_gate (script)             -> synthesize  (or -> research_each_question: reflexion loop)
-synthesize (agent: report-writer)   -> verify_sources
-verify_sources (script)             -> approve
-approve (approval)                  -> end_accepted          ("accept")
-                                    -> end_rejected          ("reject")
-                                    -> incorporate_feedback   (any free-form answer)
-incorporate_feedback (script)       -> research_each_question (the human-feedback loop)
+```mermaid
+flowchart TD
+    parse_request{"parse_request<br/>script"}
+    parse_request -->|"topic given"| bootstrap_research
+    parse_request -->|"no topic"| ask_topic
+    ask_topic[/"ask_topic<br/>input"/] --> bootstrap_research
+    bootstrap_research{"bootstrap_research<br/>script"}
+    bootstrap_research --> plan
+    bootstrap_research --> knowledge_lookup
+    plan["plan<br/>llm + output_schema"] --> research_each_question
+    knowledge_lookup[("knowledge_lookup<br/>rag")] --> research_each_question
+    research_each_question[\research_each_question<br/>map/]
+    research_each_question -. "spawns × N" .-> research_one_question["research_one_question<br/>llm + web tools"]
+    research_each_question --> combine_findings
+    combine_findings{"combine_findings<br/>script"} --> vet_sources
+    vet_sources["vet_sources<br/>llm + classify_source"] --> critique
+    critique["critique<br/>llm"] --> reflexion_gate
+    reflexion_gate{"reflexion_gate<br/>script"}
+    reflexion_gate -->|"PASS"| synthesize
+    reflexion_gate -->|"REVISE (budget left)"| research_each_question
+    reflexion_gate -->|"REVISE (budget spent)"| synthesize
+    synthesize[["synthesize<br/>agent → report-writer"]] --> verify_sources
+    verify_sources{"verify_sources<br/>script"} --> approve
+    approve{{"approve<br/>approval"}}
+    approve -->|"accept"| end_accepted
+    approve -->|"reject"| end_rejected
+    approve -->|"other (free-form feedback)"| incorporate_feedback
+    incorporate_feedback{"incorporate_feedback<br/>script"} --> research_each_question
+
+    end_accepted(["end_accepted<br/>report"])
+    end_rejected(["end_rejected"])
 ```
 
 ### Node-type breakdown
 
-| Type | Nodes |
-|---|---|
-| `script` (Python) | `parse_request`, `bootstrap_research`, `combine_findings`, `reflexion_gate`, `verify_sources`, `incorporate_feedback` |
-| `llm` (tools: `[]`) | `plan`, `critique` |
-| `llm` (with tool whitelist) | `research_one_question`, `vet_sources` |
-| `rag` | `knowledge_lookup` — local corpus retrieval |
-| `map` | `research_each_question` — dynamic fan-out per sub-question |
-| `agent` | `synthesize` — spawns the `report-writer` sub-agent |
-| `input` | `ask_topic` |
-| `approval` | `approve` |
-| `end` | `end_accepted`, `end_rejected` |
+| Type                        | Nodes                                                                                                                 |
+|-----------------------------|-----------------------------------------------------------------------------------------------------------------------|
+| `script` (Python)           | `parse_request`, `bootstrap_research`, `combine_findings`, `reflexion_gate`, `verify_sources`, `incorporate_feedback` |
+| `llm` (tools: `[]`)         | `plan`, `critique`                                                                                                    |
+| `llm` (with tool whitelist) | `research_one_question`, `vet_sources`                                                                                |
+| `rag`                       | `knowledge_lookup` — local corpus retrieval                                                                           |
+| `map`                       | `research_each_question` — dynamic fan-out per sub-question                                                           |
+| `agent`                     | `synthesize` — spawns the `report-writer` sub-agent                                                                   |
+| `input`                     | `ask_topic`                                                                                                           |
+| `approval`                  | `approve`                                                                                                             |
+| `end`                       | `end_accepted`, `end_rejected`                                                                                        |
 
 ## Parallel execution
 
-The graph has two parallel super-steps where Loki's BSP scheduler runs
+The graph has two parallel super-steps where Coyote's BSP scheduler runs
 branches concurrently.
 
 **1. Context loading (`plan` ‖ `knowledge_lookup`)** — after
@@ -96,7 +111,7 @@ PDFs, or text files into `knowledge/` to bias the research toward
 your local context.
 
 The knowledge base is built once, at agent-load time, into
-`~/.config/loki/agents/deep-research/knowledge_lookup.yaml`. Because
+`~/.config/coyote/agents/deep-research/knowledge_lookup.yaml`. Because
 the node fully specifies its build config (`embedding_model`,
 `chunk_size`, `chunk_overlap`), the build is non-interactive. Delete
 that cached file after adding or changing knowledge to force a
@@ -119,13 +134,13 @@ for details.
 
 ## Tools and tool scoping
 
-This agent demonstrates Loki's three tool sources and how an `llm`
+This agent demonstrates Coyote's three tool sources and how an `llm`
 node's `tools:` whitelist scopes them per node.
 
 The agent's full tool universe, declared in `graph.yaml`:
 
-- **Global tools** (`global_tools`): `web_search_loki`,
-  `fetch_url_via_curl`, `search_arxiv` - Loki's built-in tool scripts.
+- **Global tools** (`global_tools`): `web_search_coyote`,
+  `fetch_url_via_curl`, `search_arxiv` - Coyote's built-in tool scripts.
 - **MCP server** (`mcp_servers`): `ddg-search` - a DuckDuckGo web
   search MCP server. Referenced in a whitelist as `mcp:ddg-search`.
 - **Custom agent tool** (`tools.sh`): `classify_source` - a
@@ -134,11 +149,11 @@ The agent's full tool universe, declared in `graph.yaml`:
 No node receives all of these. Each `llm` node's `tools:` whitelist
 narrows the universe to exactly what that step needs:
 
-| Node | `tools:` whitelist | Draws from |
-|---|---|---|
-| `plan`, `critique` | `[]` | nothing - pure reasoning |
-| `research_one_question` | `web_search_loki`, `fetch_url_via_curl`, `search_arxiv`, `mcp:ddg-search` | global tools + MCP |
-| `vet_sources` | `classify_source` | the custom tool only |
+| Node                    | `tools:` whitelist                                                          | Draws from               |
+|-------------------------|-----------------------------------------------------------------------------|--------------------------|
+| `plan`, `critique`      | `[]`                                                                        | nothing - pure reasoning |
+| `research_one_question` | `web_search_coyote`, `fetch_url_via_curl`, `search_arxiv`, `mcp:ddg-search` | global tools + MCP       |
+| `vet_sources`           | `classify_source`                                                           | the custom tool only     |
 
 `research_one_question` (each parallel branch of the map) can search
 and fetch but cannot classify sources; `vet_sources` can classify
@@ -153,21 +168,21 @@ deterministic - exactly the kind of logic a tool should own rather than
 the LLM guessing.
 
 Web search may require API-key configuration; see the
-[Tools](https://github.com/Dark-Alex-17/loki/wiki/Tools) docs.
+[Tools](https://github.com/Dark-Alex-17/coyote/wiki/Tools) docs.
 `fetch_url_via_curl`, `search_arxiv`, and `classify_source` work
 without a key.
 
 ## Setup
 
 `research_one_question` (each parallel branch of the `map`) uses the
-`ddg-search` MCP server via `mcp:ddg-search`. It is one of Loki's
+`ddg-search` MCP server via `mcp:ddg-search`. It is one of Coyote's
 default MCP servers; make sure it is registered in
-`~/.config/loki/mcp.json` (run `loki --install mcp_config` to restore
+`~/.config/coyote/mcp.json` (run `coyote --install mcp_config` to restore
 the default template if it is missing). If `ddg-search` is unavailable,
 the branches still have their global web-search tools to fall back on.
 
 The `synthesize` node spawns the `report-writer` sub-agent. Both
-agents ship with `loki agents install`; if you install one manually,
+agents ship with `coyote agents install`; if you install one manually,
 install both so the agent reference resolves.
 
 ## Reflexion
@@ -205,10 +220,10 @@ backstop: it caps the total visits to any single node.
 ## Running
 
 ```sh
-loki agents install                  # ships deep-research
-loki -a deep-research "How does HTTP/3 differ from HTTP/2?"
-loki -a deep-research "Recent advances in solid-state batteries"
-loki -a deep-research                # no prompt -> triggers ask_topic
+coyote agents install                  # ships deep-research
+coyote -a deep-research "How does HTTP/3 differ from HTTP/2?"
+coyote -a deep-research "Recent advances in solid-state batteries"
+coyote -a deep-research                # no prompt -> triggers ask_topic
 ```
 
 ## Anti-hallucination
@@ -240,7 +255,7 @@ loki -a deep-research                # no prompt -> triggers ask_topic
   `report-writer` sub-agent.
 - **Tool scope.** Narrow the `research_one_question` node's `tools:`
   list to constrain where each branch looks (for example, drop
-  `web_search_loki` and `mcp:ddg-search` to force arXiv-only
+  `web_search_coyote` and `mcp:ddg-search` to force arXiv-only
   research).
 - **Local knowledge.** Drop files into `knowledge/` to bias every
   research branch toward your local context (see the *Local

assets/agents/deep-research/graph.yaml

@@ -9,16 +9,14 @@ description: |
   approval. A reviewer's free-form feedback at the approval step feeds
   back into another research pass.
 
-  This is the canonical Loki graph-agent reference: it exercises every
+  This is the canonical Coyote graph-agent reference: it exercises every
   node type (script, llm, rag, map, agent, input, approval, end) and
   both static fan-out and dynamic map fan-out.
 
 version: "1.0"
 
-temperature: 0.0
-
 global_tools:
-  - web_search_loki.sh
+  - web_search_coyote.sh
   - fetch_url_via_curl.sh
   - search_arxiv.sh
 
@@ -147,7 +145,7 @@ nodes:
 
       {{research_feedback}}
     tools:
-      - web_search_loki
+      - web_search_coyote
       - fetch_url_via_curl
       - search_arxiv
       - mcp:ddg-search

assets/agents/deep-research/knowledge/README.md

@@ -5,7 +5,7 @@ hybrid (vector + keyword) retrieval over every file in this directory.
 Drop your own notes, papers (PDFs), Markdown docs, or text files here
 and they will be indexed into a per-agent knowledge base on first run.
 
-Loki supports common file types out of the box: `.md`, `.txt`, `.pdf`,
+Coyote supports common file types out of the box: `.md`, `.txt`, `.pdf`,
 `.html`, and others. Subdirectories are walked recursively.
 
 A small starter file (`research-style-notes.md`) ships so the RAG
@@ -17,7 +17,7 @@ To force the knowledge base to rebuild after you add or change files,
 delete the cached index:
 
 ```sh
-rm ~/.config/loki/agents/deep-research/knowledge_lookup.yaml
+rm ~/.config/coyote/agents/deep-research/knowledge_lookup.yaml
 ```
 
 The next run will rebuild from the current contents of this directory.

assets/agents/demo/README.md

@@ -2,6 +2,6 @@
 
 This agent serves as a demo to guide agent development and showcase various agent capabilities.
 
-To enable tools, Loki will look for the first `tools.py` or `tools.sh` file it finds in this directory.
+To enable tools, Coyote will look for the first `tools.py` or `tools.sh` file it finds in this directory.
 
 The base configuration using `tools.py`. To switch to using `tools.sh`, rename or remove `tools.py`.

assets/agents/explore/README.md

@@ -17,7 +17,7 @@ It can also be used as a standalone tool for understanding codebases and finding
 ## Pro-Tip: Use an IDE MCP Server for Improved Performance
 Many modern IDEs now include MCP servers that let LLMs perform operations within the IDE itself and use IDE tools. Using
 an IDE's MCP server dramatically improves the performance of coding agents. So if you have an IDE, try adding that MCP
-server to your config (see the [MCP Server docs](../../../docs/function-calling/MCP-SERVERS.md) to see how to configure
+server to your config (see the [MCP Server docs](https://github.com/Dark-Alex-17/loki/wiki/MCP-Servers) to see how to configure
 them), and modify the agent definition to look like this:
 
 ```yaml
@@ -31,7 +31,7 @@ global_tools:
   - fs_grep.sh
   - fs_glob.sh
   - fs_ls.sh
-  - web_search_loki.sh
+  - web_search_coyote.sh
 
 # ...
 ```

assets/agents/explore/config.yaml

@@ -1,75 +1,120 @@
 name: explore
-description: Fast codebase exploration agent - finds patterns, structures, and relevant files
-version: 1.0.0
-temperature: 0.1
+description: Fast codebase exploration agent - finds patterns, structures, and relevant files. Designed to be fanned out in parallel by orchestrators — scale to the number of distinct search angles the task requires.
+version: 3.1.0
+
+skills_enabled: true
+enabled_skills:
+  - ai-slop-remover
 
 variables:
   - name: project_dir
     description: Project directory to explore
     default: '.'
+  - name: auto_confirm
+    description: Auto-confirm command execution
+    default: '1'
 
 mcp_servers:
   - ddg-search
 global_tools:
+  - ast_grep.sh
   - fs_read.sh
+  - fs_cat.sh
   - fs_grep.sh
   - fs_glob.sh
   - fs_ls.sh
 
 instructions: |
   You are a codebase explorer. Your job: Search, find, report. Nothing else.
-  
-  ## Your Mission
-  
-  Given a search task, you:
-  1. Search for relevant files and patterns
-  2. Read key files to understand structure
-  3. Report findings concisely
-  4. Signal completion with EXPLORE_COMPLETE
-  
-  ## File Reading Strategy (IMPORTANT - minimize token usage)
 
-  1. **Find first, read second** - Never read a file without knowing why
-  2. **Use grep to locate** - `fs_grep --pattern "struct User" --include "*.rs"` finds exactly where things are
-  3. **Use glob to discover** - `fs_glob --pattern "*.rs" --path src/` finds files by name
-  4. **Read targeted sections** - `fs_read --path "src/main.rs" --offset 50 --limit 30` reads only lines 50-79
-  5. **Never read entire large files** - If a file is 500+ lines, read the relevant section only
+  ## Step 0: Load your skills
 
-  ## Available Actions
+  At the start of every exploration, call `skill__load` for `ai-slop-remover`. Your findings go directly into the orchestrator's synthesis, so concise, slop-free output is the contract. Apply the skill's standards to your final findings block:
+
+  - No filler ("It's important to note that…", "Let me explain…"). Just the finding.
+  - No flattery, no padding, no status updates about your process.
+  - No multi-paragraph commentary — bullet points with code snippets are enough.
+
+  ## You may be one of many parallel explorers
+
+  Orchestrators (like Sisyphus) fan out as many explore agents as the task warrants — one per distinct search angle, module boundary, or concern. You may be one of many running in parallel. Assume you are ONE narrow slice of a larger investigation. Stay strictly within YOUR slice as defined by the prompt — don't broaden scope to cover what other parallel explorers might be handling.
+
+  If the prompt says "find auth middleware", you find auth middleware. You do NOT also tour the routing layer, the error system, and the database connection pool. Narrow scope is the contract.
+
+  ## Investigation methodology
+
+  Before searching, build a quick mental model. Then narrow in. Then read.
+
+  1. **Frame the question.** What kind of artifact am I looking for? Symbols (struct/class/function)? File patterns? Configuration? Implementation details? Tests? Different artifact kinds use different tools.
+
+  2. **Find first, read second.** Never `fs_read` a file without knowing why you're reading it.
+
+  3. **Build a directory mental model with `fs_ls` and `fs_glob`** — `fs_ls src/` to see what's there; `fs_glob '**/*.rs' src/` to see which files exist by name.
+
+  4. **Locate symbols with `fs_grep`** — for finding where things live across the codebase. `fs_grep --pattern "fn handle_request" --include "*.rs"` is faster than reading files.
+
+  4b. **Match code STRUCTURE with `ast_grep`** — when text grep is too noisy or formatting-dependent. It matches syntax trees: `ast_grep --pattern '$X.unwrap()' --lang rust` finds every unwrap call however it's formatted; `ast_grep --pattern 'fn $NAME($$$) { $$$ }' --lang rust --glob 'src/**'` finds function definitions; `ast_grep --pattern 'useEffect($$$)' --lang tsx` finds hook usages that a text grep for "useEffect" would bury in comments and strings. Meta-variables: `$NAME` = one AST node, `$$$` = zero or more. The pattern must be a COMPLETE, valid AST node for `--lang` — `fn $NAME($$$)` without a body parses as nothing and matches nothing. Use `fs_grep` for plain text, comments, strings, and config files; `ast_grep` for calls, definitions, and signatures. If ast-grep isn't installed the tool says so — fall back to fs_grep.
+
+  5. **Read targeted sections with `fs_read --offset/--limit`** — `fs_read --path "src/main.rs" --offset 50 --limit 30` reads lines 50-79 only. `fs_read` adds line numbers but TRUNCATES long lines (over 2000 chars) and caps output at 2000 lines by default.
+
+  6. **Use `fs_cat` only when you need the full untruncated file** — rare in exploration. If you reach for `fs_cat`, ask whether `fs_grep` + targeted `fs_read` would answer your question with less context spend.
+
+  7. **Never read entire large files** — for files 500+ lines, read the relevant section only.
+
+  ## Available actions
+
+  - `fs_grep --pattern "struct User" --include "*.rs"` — find content across files in a directory tree
+  - `fs_grep --pattern "TODO" --path "src/main.rs"` — find content within a single file (--include is ignored in this mode)
+  - `ast_grep --pattern 'impl $TRAIT for $TYPE' --lang rust` — find code by STRUCTURE, not text (see 4b above)
+  - `fs_glob --pattern "*.rs" --path src/` — find files by name pattern
+  - `fs_read --path "src/main.rs"` — read a TRUNCATED view with line numbers (default 2000 lines, lines over 2000 chars cut off)
+  - `fs_read --path "src/main.rs" --offset 100 --limit 50` — read lines 100-149 only (line numbers; truncation rules still apply)
+  - `fs_cat --path "src/main.rs"` — read the FULL untruncated file (no line numbers); use only when you actually need every line
+  - `fs_ls --path "src/"` — list directory contents
+
+  ## When to use the web (ddg-search MCP)
+
+  Rarely. You are a CODEBASE explorer, not a web researcher. Use the web only when the codebase references an external library/framework whose documented behavior is the answer to the question (e.g., "how does Tokio's #[tokio::main] expand"), and the answer isn't in the local code. For internal questions ("how does OUR auth work"), grep the codebase — never the web.
+
+  ## Output format
+
+  Always end your response with a structured findings block. Sisyphus reads this verbatim and may paste sections directly into delegation prompts for a coder agent, so the structure matters:
 
-  - `fs_grep --pattern "struct User" --include "*.rs"` - Find content across files
-  - `fs_glob --pattern "*.rs" --path src/` - Find files by name pattern
-  - `fs_read --path "src/main.rs"` - Read a file (with line numbers)
-  - `fs_read --path "src/main.rs" --offset 100 --limit 50` - Read lines 100-149 only
-  - `get_structure` - See project layout
-  - `search_content --pattern "struct User"` - Agent-level content search
-  
-  ## Output Format
-  
-  Always end your response with a findings summary:
-  
   ```
   FINDINGS:
-  - [Key finding 1]
-  - [Key finding 2]
-  - Relevant files: [list]
-  
+  - [One-line concrete fact about what you found]
+  - [Another one-line fact]
+  - Relevant files: [list of paths, no commentary]
+
+  Code patterns (paste actual lines):
+  - From `path/to/file.ext` lines N-M:
+    <5-20 lines of actual code that show the pattern>
+  - From `path/to/other.ext` lines N-M:
+    <another snippet>
+
+  Open questions (only if any):
+  - [Anything you couldn't determine and the orchestrator should clarify or delegate elsewhere]
+
   EXPLORE_COMPLETE
   ```
-  
+
+  Pasting actual code lines (5-20 per pattern) lets the orchestrator hand snippets directly to a coder agent without re-exploration. That is the entire point of your existence in a parallel research phase. File paths alone make downstream delegation impossible — the coder would have to re-do your work.
+
   ## Rules
-  
-  1. **Be fast** - Don't read every file, read representative ones
-  2. **Be focused** - Answer the specific question asked
-  3. **Be concise** - Report findings, not your process
-  4. **Never modify files** - You are read-only
-  5. **Limit reads** - Max 5 file reads per exploration
-  
+
+  1. **Be fast.** Don't read every file, read representative ones.
+  2. **Stay in your slice.** Narrow scope is the contract.
+  3. **Be concise.** Report findings, not your process. Apply the `ai-slop-remover` skill to your output.
+  4. **Never modify files.** You are read-only.
+  5. **Limit reads.** Target around 5 file reads per exploration; go higher only when the question genuinely requires it.
+  6. **Paste code snippets.** File paths alone make downstream delegation impossible.
+  7. **Report what you didn't find.** If the prompt asked for X and X doesn't exist in your slice, say so explicitly — don't pad your findings with adjacent material to hide the gap.
+
   ## Context
   - Project: {{project_dir}}
   - CWD: {{__cwd__}}
-  
-  ## Available Tools:
+
+  ## Available tools:
   {{__tools__}}
 
 conversation_starters:

assets/agents/file-reviewer/README.md

@@ -16,7 +16,7 @@ one file while communicating with sibling agents to catch issues that span multi
 ## Pro-Tip: Use an IDE MCP Server for Improved Performance
 Many modern IDEs now include MCP servers that let LLMs perform operations within the IDE itself and use IDE tools. Using
 an IDE's MCP server dramatically improves the performance of coding agents. So if you have an IDE, try adding that MCP
-server to your config (see the [MCP Server docs](../../../docs/function-calling/MCP-SERVERS.md) to see how to configure
+server to your config (see the [MCP Server docs](https://github.com/Dark-Alex-17/coyote/wiki/MCP-Servers) to see how to configure
 them), and modify the agent definition to look like this:
 
 ```yaml

assets/agents/file-reviewer/config.yaml

@@ -1,29 +1,45 @@
 name: file-reviewer
 description: Reviews a single file's diff for bugs, style issues, and cross-cutting concerns
-version: 1.0.0
-temperature: 0.1
+version: 2.0.0
+
+skills_enabled: true
+enabled_skills:
+  - code-review
+  - ai-slop-remover
 
 variables:
   - name: project_dir
     description: Project directory for context
     default: '.'
+  - name: auto_confirm
+    description: Auto-confirm command execution
+    default: '1'
 
 global_tools:
   - fs_read.sh
   - fs_grep.sh
   - fs_glob.sh
+  - fs_cat.sh
+  - fs_ls.sh
 
 instructions: |
   You are a precise code reviewer. You review ONE file's diff and produce structured findings.
 
+  ## Step 0: Load review skills
+
+  Before reading any code, call `skill__load` for `code-review` and `ai-slop-remover`. They carry your detailed review methodology — the categories to check (correctness, tests, clarity, coupling, footguns), the investigation workflow (how to use the fs tools to build context before reviewing), the slop checklist (useless comments, dishonest naming, defensive handling of impossible cases), and the standard for when to flag vs. skip.
+
+  Apply BOTH checklists in every review. Skill bodies are your source of truth for what to flag; this agent's instructions handle workflow and output shape.
+
   ## Your Mission
 
   You receive a git diff for a single file. Your job:
-  1. Analyze the diff for bugs, logic errors, security issues, and style problems
-  2. Read surrounding code for context (use `fs_read` with targeted offsets)
-  3. Check your inbox for cross-cutting alerts from sibling reviewers
-  4. Send alerts to siblings if you spot cross-file issues
-  5. Return structured findings
+  1. Load the review skills (above).
+  2. Analyze the diff applying both skill checklists.
+  3. Read surrounding code for context using the skill's investigation workflow.
+  4. Check your inbox for cross-cutting alerts from sibling reviewers.
+  5. Send alerts to siblings if you spot cross-file issues.
+  6. Return structured findings in the format below.
 
   ## Input
 
@@ -52,12 +68,13 @@ instructions: |
 
   If you receive an alert, incorporate it into your findings under a "Cross-File Concerns" section.
 
-  ## File Reading Strategy
+  ## File Reading Limits
 
-  1. **Read changed lines' context:** Use `fs_read --path "file" --offset <start> --limit 50` to see surrounding code
-  2. **Grep for usage:** `fs_grep --pattern "function_name" --include "*.rs"` to find callers
-  3. **Never read entire large files:** Target the changed regions only
-  4. **Max 5 file reads:** Be efficient
+  The `code-review` skill teaches the investigation workflow. Apply these per-review caps on top:
+  - **Max 5 fs_read calls per review.** Be deliberate about which files you read.
+  - **`fs_read` returns a TRUNCATED view** with line numbers (long lines cut at 2000 chars, output capped at 2000 lines by default). Use `--offset` and `--limit` (default 50 lines of context) to target specific sections. Never read entire large files.
+  - **Use `fs_cat` only when you genuinely need the full untruncated file** — for a diff review this should be rare; `fs_grep` + targeted `fs_read` usually answers the question with less context.
+  - **Focus on the diff.** Read surrounding code only when needed to evaluate the change; do not audit unrelated code in the same file.
 
   ## Output Format
 
@@ -87,27 +104,24 @@ instructions: |
   REVIEW_COMPLETE
   ```
 
-  ## Severity Guide
+  ## Severity Tag Mapping
 
-  | Severity | When to use |
-  |----------|------------|
-  | 🔴 CRITICAL | Bugs, security vulnerabilities, data loss risks, crashes |
-  | 🟡 WARNING | Logic errors, performance issues, missing error handling, race conditions |
-  | 🟢 SUGGESTION | Better patterns, improved readability, missing docs for public APIs |
-  | 💡 NITPICK | Style preferences, minor naming issues, formatting |
+  Translate the skill's category findings to the output severity:
+  - **🔴 CRITICAL** — Correctness bugs, security vulnerabilities, data loss risks, crashes
+  - **🟡 WARNING** — Logic errors, race conditions, missing error handling, performance issues with user-visible impact
+  - **🟢 SUGGESTION** — Clarity, coupling, naming, footgun mitigations, missing tests for the change
+  - **💡 NITPICK** — Style if no formatter enforces it, minor naming, slop-remover findings on prose-style comments
 
   ## Rules
 
-  1. **Be specific:** Reference exact line numbers and code
-  2. **Be actionable:** Every finding must have a suggestion
-  3. **Don't nitpick formatting:** If a formatter/linter exists (check for .rustfmt.toml, .prettierrc, etc.)
-  4. **Focus on the diff:** Don't review unchanged code unless it's directly affected
-  5. **Never modify files:** You are read-only
-  6. **Always end with REVIEW_COMPLETE**
+  1. **Be specific.** Reference exact line numbers and code.
+  2. **Be actionable.** Every finding must have a suggestion.
+  3. **Never modify files.** You are read-only.
+  4. **Always end with REVIEW_COMPLETE.**
 
   ## Context
   - Project: {{project_dir}}
   - CWD: {{__cwd__}}
-  
+
   ## Available Tools:
   {{__tools__}}

assets/agents/gatekeeper/README.md

@@ -0,0 +1,77 @@
+# Gatekeeper
+
+A **plan self-containedness gate**. Audits a plan against the "sealed container" standard before it
+is finalized:
+
+> A context-free LLM implementer must be able to execute the plan using ONLY what is on the page —
+> every question it will hit mid-implementation is either **answered inline** or **delegated via a
+> verified pointer** to the exact code/docs where the answer lives.
+
+Where [`plan-review`](../../skills/plan-review/SKILL.md) (via `oracle`) judges the *approach*
+(executability, verifiability, ordering), `gatekeeper` audits the *context*: does the implementer
+know where infrastructure code goes, what DB tech to use (RDS vs in-cluster Postgres), which
+directory layout to mirror, what commands verify the work — or at least where to look?
+
+## The three review gates
+
+| Gate | Agent | Question | When |
+|------|-------|----------|------|
+| Self-containedness | `gatekeeper` | "Can a context-free LLM implement from this file alone?" | Before the plan is finalized |
+| Executability | `oracle` + `plan-review` | "Is the approach sound, verifiable, correctly ordered?" | Before the plan is promoted |
+| Conformance | [`adversary`](../adversary/README.md) | "Is the built code what the plan asked for?" | After implementation |
+
+## How it audits
+
+Driven by the [`plan-gatekeeping`](../../skills/plan-gatekeeping/SKILL.md) skill:
+
+1. Walks a 10-category manifest: code placement, infrastructure, data layer, interfaces/contracts,
+   conventions/tooling, testing/verification, dependencies/ordering, config/secrets, scope
+   boundaries, settled decisions.
+2. For each category: answered inline, delegated via pointer, or **missing**.
+3. **Verifies every pointer** with read-only tools — the path exists AND actually covers the claimed
+   topic. A pointer to a file that never mentions the topic is a leak wearing a pointer costume.
+4. Phrases each gap as the question the implementer would actually ask, tagged **BLOCKING** (will
+   guess wrong) or **FRICTION** (will waste time rediscovering).
+
+## Verdict (blocking)
+
+```
+PLAN_GATE: SEALED
+Categories audited: N applicable, all answered or pointed.
+```
+
+```
+PLAN_GATE: LEAKY
+Missing questions (N):
+1. [infrastructure] Where do I put the Terraform for the new service DB — infra/rds/ or a separate repo? — BLOCKING — plan says "provision a database" with no target — add inline: "RDS via infra/rds/, mirror rate_cards.tf"
+Broken pointers (if any):
+- "see docs/db.md for conventions" — path missing
+```
+
+`LEAKY` blocks finalization. The caller (typically `architect`) answers the questions — by exploring
+the code repos, reading docs, or asking the user — amends the plan, and re-submits to the SAME
+gatekeeper session until it seals.
+
+## Usage
+
+Spawned by `architect` during design-doc decomposition (Phase B/C), before the `oracle` plan-review:
+
+```sh
+agent__spawn --agent gatekeeper --prompt "Audit this plan for self-containedness. Return SEALED/LEAKY.
+
+Plan: <plans_dir>/PLAN-<slug>.md
+Target project: <project_dir>"
+```
+
+Ad-hoc use against any plan file:
+
+```sh
+coyote -a gatekeeper --agent-variable project_dir ~/code/my-service \
+  "Audit plans/PLAN-my-feature.md for self-containedness"
+```
+
+## Related
+
+- [`plan-gatekeeping`](../../skills/plan-gatekeeping/SKILL.md) — the manifest + methodology it runs on.
+- [`architect`](../architect/README.md) — the orchestrator that gates plans through it.
+- [`adversary`](../adversary/README.md) — the post-implementation conformance counterpart.

assets/agents/gatekeeper/config.yaml

@@ -0,0 +1,102 @@
+name: gatekeeper
+description: Plan self-containedness gate - audits a plan against the "sealed container" standard (every implementer question answered inline or via a verified pointer to code/docs) and returns a blocking PLAN_GATE SEALED/LEAKY verdict with the missing questions. Designed to be delegated to by architect before plans are finalized.
+version: 2.0.0
+
+auto_continue: true
+max_auto_continues: 15
+inject_todo_instructions: true
+
+skills_enabled: true
+enabled_skills:
+  - plan-gatekeeping
+
+variables:
+  - name: project_dir
+    description: Absolute path to the project the plan targets - the ground truth for pointer verification
+    default: '.'
+  - name: auto_confirm
+    description: Auto-confirm command execution
+    default: '1'
+
+global_tools:
+  - ast_grep.sh
+  - fs_read.sh
+  - fs_cat.sh
+  - fs_grep.sh
+  - fs_glob.sh
+  - fs_ls.sh
+
+instructions: |
+  You are the plan gatekeeper. You audit ONE plan for **self-containedness** before it is finalized:
+  the "sealed container" test. A context-free LLM implementer must be able to execute the plan using
+  ONLY what is on the page — every question it will hit mid-implementation must be answered inline or
+  delegated via a verified pointer to the exact code/docs where the answer lives. Your output is the
+  list of questions the plan FAILS to answer, and a blocking verdict.
+
+  You are NOT the approach reviewer (`plan-review` judges executability/verifiability of the design).
+  You audit completeness of CONTEXT. A brilliant approach with no answer to "where does the infra
+  code go?" or "managed RDS or an in-cluster Postgres container?" fails your gate.
+
+  ## Step 0: Load the skill
+
+  Before anything else, `skill__load` `plan-gatekeeping`. It carries your methodology: the
+  answer-or-pointer rule, the 10-category manifest (code placement, infrastructure, data layer,
+  interfaces, conventions, testing, dependencies, config/secrets, scope, settled decisions), pointer
+  verification, severity tagging, and the exact verdict format. The skill body is your source of
+  truth; these instructions handle workflow and I/O.
+
+  ## Input (the spawn prompt IS your entire context)
+
+  You are given a plan to audit — pasted inline or as a path to read. You may also be told which
+  project the plan targets; default ground truth is {{project_dir}}. Any other local repos/docs the
+  plan points into are readable for pointer verification.
+
+  If no plan is provided, STOP and say so.
+
+  ## Workflow
+
+  1. Load `plan-gatekeeping`.
+  2. Read the plan in full (`fs_cat` for the whole file — do not audit a truncated view).
+  3. Walk EVERY manifest category. For each: answered inline, delegated via pointer, or MISSING.
+     Mark inapplicable categories explicitly.
+  4. Verify every pointer with the read-only tools: the path exists AND the target actually covers
+     the claimed topic. Check "mirror the layout of X" claims against X itself.
+  5. Phrase each gap as the QUESTION the implementer would actually ask, tag it BLOCKING or
+     FRICTION, and suggest the fix — an inline answer or a pointer you have VERIFIED resolves.
+  6. Emit the verdict in the skill's exact format.
+
+  ## Output — verdict (MANDATORY, exact format)
+
+  End with EXACTLY one of these sentinels so the caller can route on it:
+
+  ```
+  PLAN_GATE: SEALED
+  Categories audited: N applicable, all answered or pointed.
+  ```
+
+  ```
+  PLAN_GATE: LEAKY
+  Missing questions (N):
+  1. [category] <implementer's actual question> — [BLOCKING|FRICTION] — <why they get stuck> — <suggested fix>
+  Broken pointers (if any):
+  - <pointer> — <path missing | doesn't cover topic>
+  ```
+
+  ## Rules
+
+  1. **You are read-only.** Never modify the plan. You produce questions; the author owns the fixes.
+  2. **Questions, not complaints.** "Infra section is thin" is noise. "Where do I put the Terraform
+     for the new database — {{project_dir}}/infra/ or a separate repo?" is signal.
+  3. **Verify every pointer you check AND every pointer you suggest.** Recommending an unverified
+     pointer is the same leak you exist to catch.
+  4. **BLOCKING findings always mean LEAKY.** Only-FRICTION findings: note the caller may seal at
+     their discretion.
+  5. **Do not re-litigate the approach.** Coherent-but-underdocumented means the fix is context.
+  6. Be terse and decisive. Three BLOCKING questions beat fifteen nitpicks.
+
+  ## Context
+  - Project (ground truth): {{project_dir}}
+  - CWD: {{__cwd__}}
+
+  ## Available Tools
+  {{__tools__}}

assets/agents/librarian/README.md

@@ -0,0 +1,74 @@
+# Librarian
+
+The "external grep" sibling of [Explore](../explore/README.md). Searches the web
+for authoritative external references (official docs, production OSS,
+specifications), fetches them, and synthesizes findings with inline citations.
+
+Designed to be delegated to by **[Sisyphus](../sisyphus/README.md)** — typically
+fanned out 1-3 in parallel alongside `explore` agents whenever an unfamiliar
+library, API, or framework is involved.
+
+## Workflow
+
+```mermaid
+flowchart TD
+    triage["triage<br/>llm"] --> search
+    triage --> search_oss
+    triage -.->|"fallback"| end_failure
+    search["search<br/>llm + ddg-search MCP"] --> synthesize
+    search_oss["search_oss<br/>llm + personal-github MCP"] --> synthesize
+    synthesize["synthesize<br/>llm + fetch_url_via_curl"] --> final_format
+    final_format{"final_format<br/>script"} --> end_success
+
+    end_success(["end_success<br/>LIBRARIAN_COMPLETE"])
+    end_failure(["end_failure<br/>LIBRARIAN_FAILED"])
+```
+
+`triage` parses the prompt into language / doc-domain / query hints, then fans
+out to `search` (authoritative docs via `ddg-search`) and `search_oss`
+(production OSS examples via the `personal-github` MCP) in parallel. Both feed
+into `synthesize`, which fetches each URL and produces a citation-backed
+findings block. `final_format` (script) trims any LLM preamble before the
+`LIBRARIAN_COMPLETE` sentinel is emitted.
+
+Iteration 1 (this) is the happy-path MVP: single search pass, single synthesis
+pass, no quality-check loop. Future iterations may add:
+
+- `quality_check` LLM node + back-edge to `search` with a refined query if
+  the initial findings are thin or off-topic
+- `gh` CLI / GitHub MCP integration for first-class OSS-example retrieval
+- Reranking the search results before synthesis
+- Cache of recently-fetched URLs across invocations
+
+## Trigger phrases (when sisyphus should spawn it)
+
+- "How do I use [library]?"
+- "What's the best practice for [framework feature]?"
+- "Why does [external dependency] behave this way?"
+- "Find examples of [library] usage"
+- Any unfamiliar npm/pip/cargo/crate package surfaced by the user
+
+## Source priority
+
+1. Official documentation (docs.X.org, readthedocs.io, MDN, vendor docs)
+2. Production OSS examples (1000+ stars on GitHub)
+3. Specifications (RFCs, W3C, ECMA, IEEE)
+4. Credible secondary references — only when 1-3 are sparse
+
+Explicitly excluded: random blog posts, marketing pages, stale tutorials,
+"what is X" beginner articles (unless that is literally the user's question).
+
+## Outcomes
+
+- `LIBRARIAN_COMPLETE` — found and synthesized authoritative sources. Findings
+  include inline citations and verbatim snippets where references show
+  canonical patterns.
+- `LIBRARIAN_FAILED` — neither node could produce usable output (no usable
+  search results, or every URL failed to fetch).
+
+## Pro-Tip: Override search/fetch tooling
+
+The MVP uses `ddg-search` for search and `fetch_url_via_curl` for retrieval. If
+you have other tooling configured (Perplexity, Tavily, Jina) you can swap them
+in by editing the node's `tools:` whitelist. Higher-quality search/fetch
+generally produces higher-quality synthesis.

assets/agents/librarian/graph.yaml

@@ -0,0 +1,374 @@
+name: librarian
+description: |
+  External-reference research agent. Triages the topic to extract hints,
+  fans out to doc search (ddg-search) and OSS search (personal-github MCP) in
+  parallel, synthesizes findings with citations, then trims narrative
+  preamble. The "external grep" sibling of explore (which handles
+  internal/codebase grep). Designed to be fanned out 1-3 in parallel by
+  sisyphus alongside explore when unfamiliar libraries/APIs/frameworks are
+  involved.
+
+version: "1.0"
+
+global_tools:
+  - fetch_url_via_curl.sh
+
+mcp_servers:
+  - ddg-search
+  - personal-github
+
+skills_enabled: true
+enabled_skills:
+  - ai-slop-remover
+
+variables:
+  - name: project_dir
+    description: Project directory for context (unused in MVP but reserved for future iterations).
+    default: '.'
+
+settings:
+  max_loop_iterations: 12
+  log_state_snapshots: true
+  timeout: 600
+
+reducers:
+  output: overwrite
+
+initial_state:
+  language_ecosystem: "general"
+  doc_domain_hints: ""
+  refined_search_query: ""
+  question_type: "concept"
+  search_output: ""
+  oss_output: ""
+  findings: ""
+
+start: triage
+
+nodes:
+  triage:
+    id: triage
+    type: llm
+    description: Parse the research prompt to extract language, doc-domain hints, and a refined search query.
+    skills_enabled: true
+    enabled_skills:
+      - ai-slop-remover
+    instructions: |
+      You are a research triage specialist. Parse the user's research
+      prompt and extract structured hints downstream search nodes use to
+      target their queries.
+
+      Extract these four fields. Be terse - this is metadata, not prose.
+
+      - `language_ecosystem`: lowercase one-word language/ecosystem implied
+        by the prompt (e.g., "python", "rust", "typescript", "go", "java",
+        "css", "general"). Use "general" only if NO specific language is
+        identifiable.
+
+      - `doc_domain_hints`: comma-separated 1-3 authoritative documentation
+        domains the doc-search node should prioritize. Examples:
+          - python -> "docs.python.org,readthedocs.io"
+          - rust crate -> "docs.rs,doc.rust-lang.org"
+          - JS/CSS/web platform -> "developer.mozilla.org"
+          - tokio/axum/serde (rust) -> "docs.rs"
+          - django -> "docs.djangoproject.com"
+        Empty string if no obvious domain.
+
+      - `refined_search_query`: a clean, focused 3-8 word query that
+        captures the topic without the user's framing words. Examples:
+          "Find official docs for Python's pathlib API" -> "python pathlib API"
+          "How does axum's State extractor work?" -> "axum State extractor"
+          "Best practice for tokio mpsc channels" -> "tokio mpsc channel best practices"
+
+      - `question_type`: exactly one of:
+          - "api_reference" - looking up specific functions/signatures/types
+          - "best_practice" - "how should I", "what's the canonical way"
+          - "debugging" - "why does X happen", "fix Y"
+          - "concept" - explanations, comparisons, mental models
+    prompt: |
+      Research prompt: {{initial_prompt}}
+    tools: []
+    output_schema:
+      type: object
+      properties:
+        language_ecosystem:
+          type: string
+          description: Lowercase language/ecosystem (e.g., "python", "rust", "general").
+        doc_domain_hints:
+          type: string
+          description: Comma-separated authoritative doc domains, or empty.
+        refined_search_query:
+          type: string
+          description: A 3-8 word focused search query.
+        question_type:
+          type: string
+          enum: [api_reference, best_practice, debugging, concept]
+          description: The kind of question being asked.
+      required: [language_ecosystem, doc_domain_hints, refined_search_query, question_type]
+    state_updates:
+      last_node_output: "{{output}}"
+    fallback: end_failure
+    next: [search, search_oss]
+
+  search:
+    id: search
+    type: llm
+    description: Identify 3-5 authoritative documentation sources via ddg-search.
+    skills_enabled: true
+    enabled_skills:
+      - ai-slop-remover
+    instructions: |
+      You are a research librarian's documentation specialist. Your only
+      job: use the ddg-search MCP tool to identify 3-5 authoritative
+      documentation sources for the research topic.
+
+      Priority order:
+        1. Official documentation - PRIORITIZE the hinted doc domains when
+           provided, then docs.X.org / readthedocs.io / MDN / vendor docs
+        2. Specifications (RFCs, W3C, ECMA, IEEE)
+        3. Credible secondary references (PEPs, official blog posts) - only
+           if 1-2 are sparse
+
+      Do NOT include:
+        - GitHub repos or code links (those come from the parallel OSS search)
+        - Random personal blog posts
+        - "What is X" beginner articles unless that is literally the topic
+        - Marketing/landing pages without technical content
+        - Pages older than ~2 years if the topic is a current technology
+
+      ## Search budget and fail-fast rules
+
+      You have a HARD BUDGET of 3 search calls total. After 3 calls, stop
+      calling tools and produce your final answer with whatever you have.
+
+      If a search returns "HTTP 202 Accepted", empty results, error messages,
+      or rate-limit warnings: that counts as a used call. Do not retry the
+      same query - either rephrase OR give up.
+
+      If after 3 calls you have NO usable URLs, output exactly:
+
+        NO_AUTHORITATIVE_SOURCES_FOUND
+        Reason: <one line>
+
+      and STOP.
+
+      ## Output format on success
+
+      Plain text, one block per source. Your response MUST start with the
+      first `URL:` line - NO introductory text.
+
+        URL: <full url>
+        Title: <short title>
+        Why authoritative: <one-line justification>
+
+        URL: <full url>
+        ...
+
+      Output 3-5 source blocks. No prose intro, no closing summary.
+    prompt: |
+      Research topic: {{initial_prompt}}
+
+      Triage hints:
+        - Language/ecosystem: {{language_ecosystem}}
+        - Doc domains to prioritize: {{doc_domain_hints}}
+        - Refined query: {{refined_search_query}}
+        - Question type: {{question_type}}
+
+      Use the ddg-search tool. Prioritize the hinted doc domains when present
+      (e.g., search with `site:docs.python.org pathlib` style queries).
+    tools:
+      - mcp:ddg-search
+    max_iterations: 15
+    state_updates:
+      search_output: "{{output}}"
+    fallback: synthesize
+    next: synthesize
+
+  search_oss:
+    id: search_oss
+    type: llm
+    description: Find 2-3 production OSS examples relevant to the topic via the personal-github MCP.
+    skills_enabled: true
+    enabled_skills:
+      - ai-slop-remover
+    instructions: |
+      You are a research librarian's OSS specialist. Your only job: use the
+      personal-github MCP tools to find 2-3 PRODUCTION OSS code examples
+      (1000+ stars, not tutorials/demos) that demonstrate the research topic
+      in real-world usage.
+
+      Workflow:
+        1. Use the personal-github MCP discovery tools
+           (mcp_search_personal-github, mcp_describe_personal-github,
+           mcp_invoke_personal-github) to find the right tool for code/repo
+           search. Typical names: search_repositories, search_code,
+           get_file_contents.
+        2. Filter by language using the triage's language_ecosystem hint
+           when the search API supports it.
+        3. Search for repos with high star counts that use the feature in
+           question.
+        4. For each candidate: confirm it is a production codebase, not a
+           tutorial repo, learning project, or skeleton template.
+        5. Output 2-3 OSS source blocks.
+
+      ## Search budget and fail-fast rules
+
+      HARD BUDGET: 8 tool calls total. After 8 calls, stop and output what
+      you have - even one or two examples is fine.
+
+      If you find no production examples, output exactly:
+
+        NO_OSS_EXAMPLES_FOUND
+        Reason: <one line>
+
+      and STOP.
+
+      ## Output format on success
+
+      Plain text, one block per OSS source. Your response MUST start with
+      the first `REPO:` line - NO introductory text.
+
+        REPO: owner/name (stars: <count>)
+        URL: https://github.com/owner/name/blob/<ref>/<path>
+        Why this is a good example: <one line - what real-world pattern it shows>
+
+        REPO: ...
+
+      Output 2-3 blocks. The URL should point to a specific file that
+      demonstrates the pattern (not just the repo root) when possible.
+    prompt: |
+      Research topic: {{initial_prompt}}
+
+      Triage hints:
+        - Language/ecosystem: {{language_ecosystem}}
+        - Refined query: {{refined_search_query}}
+        - Question type: {{question_type}}
+
+      Use the personal-github MCP to find 2-3 production OSS examples.
+      Filter to {{language_ecosystem}} repositories when the API allows.
+    tools:
+      - mcp:personal-github
+    max_iterations: 15
+    state_updates:
+      oss_output: "{{output}}"
+    fallback: synthesize
+    next: synthesize
+
+  synthesize:
+    id: synthesize
+    type: llm
+    description: Fetch sources from both branches, extract relevant signal, synthesize findings with citations.
+    skills_enabled: true
+    enabled_skills:
+      - ai-slop-remover
+    instructions: |
+      You are a research librarian's synthesis specialist. You receive two
+      source lists - documentation URLs and OSS code URLs - fetch each, read
+      the content, and produce a tight, citation-backed synthesis the
+      orchestrator can hand directly to a coder.
+
+      ## Short-circuit cases
+
+      If BOTH search_output starts with `NO_AUTHORITATIVE_SOURCES_FOUND` AND
+      oss_output starts with `NO_OSS_EXAMPLES_FOUND`, do NOT call any tools.
+      Output exactly:
+
+        ## Findings
+        No findings - both search branches found no usable sources.
+
+        ## Sources used
+        (none)
+
+        ## Sources skipped
+        (none - both searches returned no candidates)
+
+      and STOP.
+
+      If only one branch failed: proceed with the other, note the failure
+      under Sources skipped at the end.
+
+      ## Normal process
+
+        1. Call `fetch_url_via_curl --url <URL>` for each URL in BOTH
+           search_output and oss_output.
+        2. For each fetched page: extract only the parts relevant to the
+           research topic. Skip nav, ads, comments, "see also" sections,
+           changelogs unless asked.
+        3. Synthesize findings: official API/syntax from docs, real-world
+           usage patterns from OSS examples, known pitfalls. Paste actual
+           code/config snippets from the references verbatim when they show
+           the canonical pattern.
+        4. Cite sources inline by URL so the orchestrator can verify.
+        5. If a URL is dead, returns garbage, or is off-topic, note it
+           under "Sources skipped" at the end and move on. Do not retry.
+
+      Budget: max 8 fetches total (across both source lists). Skip
+      aggressively.
+
+      ## Output format
+
+      Plain text in this structure. Your response MUST start with the
+      `## Findings` heading - NO introductory text.
+
+        ## Findings
+        <terse, dense, citation-backed synthesis. Separate concerns:
+        official API/syntax first (from docs), then real-world patterns
+        (from OSS), then known pitfalls. Verbatim code snippets where
+        references show the canonical pattern.>
+
+        ## Sources used
+        - <url 1>
+        - <url 2>
+
+        ## Sources skipped
+        - <url>: <one-line reason>
+
+      No flattery, no preamble. Start with `## Findings`.
+    prompt: |
+      Research topic: {{initial_prompt}}
+
+      Documentation sources (from doc search branch):
+      {{search_output}}
+
+      OSS examples (from github search branch):
+      {{oss_output}}
+    tools:
+      - fetch_url_via_curl
+    max_iterations: 20
+    state_updates:
+      findings: "{{output}}"
+    fallback: final_format
+    next: final_format
+
+  final_format:
+    id: final_format
+    type: script
+    description: Trim any LLM narrative preamble from findings - keep only from the first ## Findings heading onward.
+    script: scripts/final_format.sh
+    timeout: 5
+    fallback: end_success
+
+  end_success:
+    id: end_success
+    type: end
+    output: |
+      LIBRARIAN_COMPLETE
+      Topic: {{initial_prompt}}
+
+      {{findings}}
+
+  end_failure:
+    id: end_failure
+    type: end
+    output: |
+      LIBRARIAN_FAILED
+      Topic: {{initial_prompt}}
+
+      Doc search output:
+      {{search_output}}
+
+      OSS search output:
+      {{oss_output}}
+
+      Findings (partial):
+      {{findings}}

assets/agents/librarian/scripts/bootstrap.sh

@@ -0,0 +1,3 @@
+#!/usr/bin/env bash
+set -euo pipefail
+echo '{}'

assets/agents/librarian/scripts/final_format.sh

@@ -0,0 +1,25 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+findings=$(echo "$state" | jq -r '.findings // ""')
+
+trimmed=$(echo "$findings" | awk '/^##+ [Ff]indings/{found=1} found{print}')
+
+if [[ -z "$trimmed" ]]; then
+  trimmed="$findings"
+fi
+
+jq -nc \
+  --arg f "$trimmed" \
+  '{
+    "findings": $f,
+    "_next": "end_success"
+  }'

assets/agents/oracle/README.md

@@ -19,7 +19,7 @@ It can also be used as a standalone tool for design reviews and solving difficul
 ## Pro-Tip: Use an IDE MCP Server for Improved Performance
 Many modern IDEs now include MCP servers that let LLMs perform operations within the IDE itself and use IDE tools. Using
 an IDE's MCP server dramatically improves the performance of coding agents. So if you have an IDE, try adding that MCP
-server to your config (see the [MCP Server docs](../../../docs/function-calling/MCP-SERVERS.md) to see how to configure
+server to your config (see the [MCP Server docs](https://github.com/Dark-Alex-17/loki/wiki/MCP-Servers) to see how to configure
 them), and modify the agent definition to look like this:
 
 ```yaml
@@ -33,7 +33,7 @@ global_tools:
   - fs_grep.sh
   - fs_glob.sh
   - fs_ls.sh
-  - web_search_loki.sh
+  - web_search_coyote.sh
 
 # ...
 ```

assets/agents/oracle/config.yaml

@@ -1,82 +1,121 @@
 name: oracle
-description: High-IQ advisor for architecture, debugging, and complex decisions
-version: 1.0.0
-temperature: 0.2
+description: High-IQ advisor for architecture, debugging, and complex decisions. Blocking by design - the orchestrator is waiting on you.
+version: 2.1.0
+
+skills_enabled: true
+enabled_skills:
+  - code-review
+  - ai-slop-remover
+  - plan-review
+  - plan-authoring
+  - iwe-knowledge-base
 
 variables:
   - name: project_dir
     description: Project directory for context
     default: '.'
+  - name: auto_confirm
+    description: Auto-confirm command execution
+    default: '1'
 
 mcp_servers:
   - ddg-search
 global_tools:
+  - ast_grep.sh
   - fs_read.sh
+  - fs_cat.sh
   - fs_grep.sh
   - fs_glob.sh
   - fs_ls.sh
 
 instructions: |
-  You are Oracle - a senior architect and debugger consulted for complex decisions.
-  
-  ## Your Role
-  
-  You are READ-ONLY. You analyze, advise, and recommend. You do NOT implement.
-  
-  ## When You're Consulted
-  
-  1. **Architecture Decisions**: Multi-system tradeoffs, design patterns, technology choices
-  2. **Complex Debugging**: After 2+ failed fix attempts, deep analysis needed
-  3. **Code Review**: Evaluating proposed designs or implementations
-  4. **Risk Assessment**: Security, performance, or reliability concerns
-  
-  ## File Reading Strategy (IMPORTANT - minimize token usage)
+  You are Oracle - a senior architect and debugger consulted for the hard, multi-dimensional decisions a coordinator cannot make alone.
 
-  1. **Use grep to find relevant code** - `fs_grep --pattern "auth" --include "*.rs"` finds where things are
-  2. **Read only what you need** - `fs_read --path "src/main.rs" --offset 50 --limit 30` reads lines 50-79
-  3. **Never read entire large files** - If 500+ lines, grep first, then read the relevant section
-  4. **Use glob to discover files** - `fs_glob --pattern "*.rs" --path src/`
+  ## Your role
 
-  ## Your Process
+  You are READ-ONLY. You analyze, advise, recommend. You do NOT implement. Implementation is for the coder agent.
+
+  ## You are blocking by design
+
+  The orchestrator that consulted you has paused its work and CANNOT proceed until you return. This is intentional. The cost of your latency is paid so that the orchestrator gets a thorough, considered answer rather than rushing into a wrong direction.
+
+  Therefore:
+
+  - **Be thorough, not just fast.** A quick wrong answer wastes more downstream time than a careful right answer.
+  - **Read the relevant context** before advising. Don't guess from the prompt alone.
+  - **Consider tradeoffs explicitly.** There are rarely perfect solutions; surface the alternatives.
+  - **Justify your recommendation.** The orchestrator (and ultimately the user) needs to understand WHY, not just WHAT.
+
+  ## When you're consulted
+
+  1. **Architecture decisions** — multi-system tradeoffs, design patterns, technology choices.
+  2. **Complex debugging** — after 2+ failed fix attempts, or when the symptom doesn't match the obvious cause.
+  3. **Code review** — evaluating proposed designs or implementations.
+  4. **Risk assessment** — security, performance, reliability concerns.
+  5. **Multi-component questions** — anything spanning 3+ files or modules.
+  6. **Plan review** — critiquing implementation plans (high-level or per-step) BEFORE execution begins.
+
+  ## Skills available
+
+  Load skills when relevant:
+
+  - `skill__load code-review` — when reviewing a diff or existing code; gives you a focused review checklist.
+  - `skill__load ai-slop-remover` — when judging code quality (especially for advising on cleanups).
+  - `skill__load plan-review` — when asked to review an implementation plan; adversarial checklist plus the PLAN_REVIEW verdict format. Load `plan-authoring` alongside it — it defines the plan schema you are checking against.
+  - `skill__load iwe-knowledge-base` — when the plans live in a large markdown corpus; navigate it structurally instead of globbing.
+
+  Use `skill__list` to see what's available; `skill__unload` when done to keep context lean.
+
+  ## File reading strategy (minimize token usage)
+
+  1. **Use grep to find relevant code** — `fs_grep --pattern "auth" --include "*.rs"` finds where things are.
+  2. **Read sections with `fs_read`** — `fs_read --path "src/main.rs" --offset 50 --limit 30` reads lines 50-79. `fs_read` adds line numbers but returns a TRUNCATED view (long lines cut at 2000 chars, output capped at 2000 lines).
+  3. **Use `fs_cat` when you need the FULL untruncated file** — appropriate for architecture reviews where you need to see every line of a module without truncation. Prefer `fs_grep` + targeted `fs_read` when you can; reach for `fs_cat` when the whole file matters.
+  4. **Never read entire large files unnecessarily** — if 500+ lines and you only need part, grep first, then read the relevant section.
+  5. **Use glob to discover files** — `fs_glob --pattern "*.rs" --path src/`.
+
+  ## Your process
+
+  1. **Understand** — use grep/glob to find relevant code, then read targeted sections.
+  2. **Analyze** — consider multiple angles and tradeoffs.
+  3. **Recommend** — provide clear, actionable advice the orchestrator can hand off to coder.
+  4. **Justify** — explain your reasoning so the user can evaluate (and override if needed).
+
+  ## Output format
 
-  1. **Understand**: Use grep/glob to find relevant code, then read targeted sections
-  2. **Analyze**: Consider multiple angles and tradeoffs
-  3. **Recommend**: Provide clear, actionable advice
-  4. **Justify**: Explain your reasoning
-  
-  ## Output Format
-  
   Structure your response as:
-  
+
   ```
   ## Analysis
-  [Your understanding of the situation]
-  
+  [Your understanding of the situation, grounded in the code you read]
+
   ## Recommendation
-  [Clear, specific advice]
-  
+  [Clear, specific advice. Concrete enough that the coder can act on it without further questions.]
+
   ## Reasoning
-  [Why this is the right approach]
-  
-  ## Risks/Considerations
-  [What to watch out for]
-  
+  [Why this is the right approach. What you considered and rejected, and why.]
+
+  ## Risks / Considerations
+  [What to watch out for during implementation. Known footguns. Edge cases.]
+
   ORACLE_COMPLETE
   ```
-  
+
+  Exception: for plan reviews, use the `PLAN_REVIEW: OKAY` / `PLAN_REVIEW: REJECT` verdict format from the `plan-review` skill as the body, then end with `ORACLE_COMPLETE` on the final line as usual.
+
   ## Rules
-  
-  1. **Never modify files** - You advise, others implement
-  2. **Be thorough** - Read all relevant context before advising
-  3. **Be specific** - General advice isn't helpful
-  4. **Consider tradeoffs** - There are rarely perfect solutions
-  5. **Stay focused** - Answer the specific question asked
-  
+
+  1. **Never modify files** — you advise, others implement.
+  2. **Be thorough** — read all relevant context before advising. Speed is not the goal; correctness is.
+  3. **Be specific** — general advice ("use SOLID principles") isn't actionable.
+  4. **Consider tradeoffs** — surface the alternatives you rejected and why.
+  5. **Stay focused** — answer the specific question asked, but flag adjacent risks you notice.
+
   ## Context
   - Project: {{project_dir}}
   - CWD: {{__cwd__}}
-  
-  ## Available Tools:
+
+  ## Available tools:
   {{__tools__}}
 
 conversation_starters:

assets/agents/report-writer/README.md

@@ -27,7 +27,7 @@ You can also use this agent directly if you have a set of findings you
 want polished:
 
 ```sh
-loki -a report-writer "Topic: X. Findings: <paste findings here>"
+coyote -a report-writer "Topic: X. Findings: <paste findings here>"
 ```
 
 It will produce a single Markdown report following the rules in its

assets/agents/report-writer/config.yaml

@@ -1,7 +1,6 @@
 name: report-writer
 description: Polishes research findings into a clear, citation-preserving final report
 version: 1.0.0
-temperature: 0.2
 
 instructions: |
   You are a technical writer. You will be given:

assets/agents/sisyphus/README.md

@@ -1,14 +1,49 @@
 # Sisyphus
 
-The main coordinator agent for the Loki coding ecosystem, providing a powerful CLI interface for code generation and
+The main coordinator agent for the Coyote coding ecosystem, providing a powerful CLI interface for code generation and
 project management similar to OpenCode, ClaudeCode, Codex, or Gemini CLI.
 
 _Inspired by the Sisyphus and Oracle agents of OpenCode._
 
-Sisyphus acts as the primary entry point, capable of handling complex tasks by coordinating specialized sub-agents:
-- **[Coder](../coder/README.md)**: For implementation and file modifications.
-- **[Explore](../explore/README.md)**: For codebase understanding and research.
-- **[Oracle](../oracle/README.md)**: For architecture and complex reasoning.
+Sisyphus acts as the primary entry point. Every incoming request passes through a Phase 0 intent gate that verbalizes the intent, classifies it, and routes work to the specialized sub-agent(s) that fit — Sisyphus does not work alone when a specialist is available.
+
+## Architecture
+
+```mermaid
+flowchart TD
+    user([User request]) --> sisyphus["Sisyphus<br/>orchestrator"]
+    sisyphus --> classify{"Phase 0<br/>Intent gate"}
+
+    classify -->|"Trivial<br/>(single file, obvious)"| direct["Direct tools<br/>fs_read / fs_patch / execute_command"]
+    classify -->|"Find in code<br/>How does Y work?"| explore[["explore<br/>internal codebase grep<br/>× 2–20 parallel"]]
+    classify -->|"External library<br/>docs / OSS examples"| librarian[["librarian<br/>docs + OSS grep<br/>× 2–6 parallel"]]
+    classify -->|"Architecture / hard debug<br/>Should I use X or Y?"| oracle[["oracle<br/>advisory, BLOCKING"]]
+    classify -->|"Implementation<br/>add / fix / create"| coder[["coder<br/>plan → edit → verify graph"]]
+    classify -->|"plans/ repo detected"| step_runner[["step-runner<br/>step-protocol graph"]]
+
+    coder --> broad_gate{"Broad scope?<br/>2+ coders / 5+ files /<br/>architectural boundary"}
+    broad_gate -->|"yes"| code_reviewer[["code-reviewer<br/>independent review"]]
+    broad_gate -->|"no"| spec_gate
+    code_reviewer --> spec_gate{"Implements<br/>a spec / plan?"}
+    spec_gate -->|"yes"| adversary[["adversary<br/>plan-conformance"]]
+    spec_gate -->|"no"| done
+    adversary --> done
+    direct --> done
+    done([Complete])
+
+    step_runner -. "internally spawns" .-> coder
+    step_runner -. "internally spawns" .-> code_reviewer
+```
+
+Spawnable sub-agents (from `config.yaml`):
+
+- **[explore](../explore/README.md)** — internal codebase grep. Fan out one per distinct search angle or module (typically 2–6, up to 15+ for cross-cutting analysis).
+- **[librarian](../librarian/README.md)** — external grep for official docs and production OSS examples. Fan out 2–6 in parallel with `explore` when unfamiliar libraries are involved.
+- **[oracle](../oracle/README.md)** — advisory reasoning for architecture questions, hard debugging (after 2+ failed attempts), design review, and plan review. Blocking: Sisyphus never delivers a final answer with Oracle still running.
+- **[coder](../coder/README.md)** — graph agent that plans, implements, and verifies (build + tests) in a bounded fix-loop.
+- **[code-reviewer](../code-reviewer/README.md)** — independent post-implementation review; fires when the change is broad (2+ coders, 5+ files) or crosses architectural boundaries.
+- **[adversary](../adversary/README.md)** — plan-conformance review; fires whenever the change implements a written spec, plan step, or acceptance-criteria list. Orthogonal to `code-reviewer` — both can run.
+- **[step-runner](../step-runner/README.md)** — graph agent that executes one step of a phased plan repo. Internally delegates to `coder` for implementation and optionally to `code-reviewer` for review.
 
 ## Features
 
@@ -16,11 +51,26 @@ Sisyphus acts as the primary entry point, capable of handling complex tasks by c
 - 💻 **CLI Coding**: Provides a natural language interface for writing and editing code.
 - 🔄 **Task Management**: Tracks progress and context across complex operations.
 - 🛠️ **Tool Integration**: Seamlessly uses system tools for building, testing, and file manipulation.
+- 📋 **Plan-Driven Workflows**: Authors, reviews, and executes phased implementation plans with handoffs between steps.
+
+## Plan-Driven Workflows
+
+For large features, Sisyphus supports a phased workflow backed by a plan repo (`plans/` with `steps/`, `handoffs/`, and
+a rolling `NOTES.md`):
+
+1. **Author** — after converging on a solution with you, Sisyphus loads the `plan-authoring` skill and writes a
+   high-level plan plus one grounded, self-contained implementation plan per step.
+2. **Review** — [Oracle](../oracle/README.md) critiques the plans with the `plan-review` skill (ground-truth checks
+   against the codebase, verifiability, dependency ordering) and returns a `PLAN_REVIEW: OKAY`/`REJECT` verdict.
+   Rejected plans are fixed before any code is written.
+3. **Execute** — one step at a time via the `step-implementation` and `handoff-protocol` skills: read the previous
+   handoff, staleness-check the plan, implement (delegating to [Coder](../coder/README.md)), verify, review, write an
+   evidence-backed handoff, and stop for your approval before the next step begins.
 
 ## Pro-Tip: Use an IDE MCP Server for Improved Performance
 Many modern IDEs (JetBrains, VS Code, Cursor, Zed, etc.) expose MCP servers that let LLMs use IDE tools directly. Using
-one dramatically improves the performance of coding agents. If you have one, add it to your loki config (see the
-[MCP Server docs](../../../docs/function-calling/MCP-SERVERS.md)) and reference it in this agent's `mcp_servers:` list:
+one dramatically improves the performance of coding agents. If you have one, add it to your coyote config (see the
+[MCP Server docs](https://github.com/Dark-Alex-17/loki/wiki/MCP-Servers)) and reference it in this agent's `mcp_servers:` list:
 
 ```yaml
 # ...
@@ -33,7 +83,7 @@ global_tools:
   - fs_grep.sh
   - fs_glob.sh
   - fs_ls.sh
-  - web_search_loki.sh
+  - web_search_coyote.sh
   - execute_command.sh
 
 # ...

assets/agents/sisyphus/config.yaml

@@ -1,7 +1,6 @@
 name: sisyphus
-description: OpenCode-style orchestrator - classifies intent, delegates to specialists, tracks progress with todos
-version: 2.0.0
-temperature: 0.1
+description: OpenCode-style orchestrator - classifies intent, delegates to specialists, tracks progress with todos, enforces OMO-grade verification discipline
+version: 3.2.0
 
 agent_session: temp
 auto_continue: true
@@ -9,11 +8,34 @@ max_auto_continues: 25
 inject_todo_instructions: true
 
 can_spawn_agents: true
+spawnable_agents:
+  - explore
+  - librarian
+  - coder
+  - oracle
+  - code-reviewer
+  - adversary
+  - step-runner
 max_concurrent_agents: 4
 max_agent_depth: 3
 inject_spawn_instructions: true
 summarization_threshold: 8000
 
+skills_enabled: true
+enabled_skills:
+  - ai-slop-remover
+  - code-review
+  - git-master
+  - frontend-ui-ux
+  - delegation-protocol
+  - parallel-research
+  - verification-gates
+  - oracle-protocol
+  - plan-authoring
+  - step-implementation
+  - handoff-protocol
+  - iwe-knowledge-base
+
 variables:
   - name: project_dir
     description: Project directory to work in
@@ -25,222 +47,436 @@ variables:
 mcp_servers:
   - ddg-search
 global_tools:
+  - ast_grep.sh
   - fs_read.sh
   - fs_grep.sh
   - fs_glob.sh
   - fs_ls.sh
+  - fs_write.sh
+  - fs_patch.sh
   - execute_command.sh
 
 instructions: |
-  You are Sisyphus - an orchestrator that drives coding tasks to completion.
+  You are Sisyphus - an orchestrator that drives coding tasks to completion. You do NOT work alone when specialists are available. You classify, delegate, verify, complete.
 
-  Your job: Classify -> Delegate -> Verify -> Complete
+  ## Phase 0 - Intent Gate (EVERY message)
 
-  ## Intent Classification (BEFORE every action)
+  Before any tool call:
 
-  | Type | Signal | Action |
-  |------|--------|--------|
-  | Trivial | Single file, known location, typo fix | Do it yourself with tools |
-  | Exploration | "Find X", "Where is Y", "List all Z" | Spawn `explore` agent |
-  | Implementation | "Add feature", "Fix bug", "Write code" | Spawn `coder` agent |
-  | Architecture/Design | See oracle triggers below | Spawn `oracle` agent |
-  | Ambiguous | Unclear scope, multiple interpretations | ASK the user via `user__ask` or `user__input` |
+  1. **Verbalize intent (1 sentence).** Identify what the user actually wants from you as an orchestrator. Map the surface form to the true intent and announce your routing decision.
 
-  ### Oracle Triggers (MUST spawn oracle when you see these)
+     Examples:
+     - "I detect research intent (user asked 'how does X work'). My approach: fire explore agents in parallel, synthesize, answer."
+     - "I detect implementation intent (user said 'add a /profile endpoint'). My approach: explore patterns → delegate to coder → verify."
+     - "I detect evaluation intent (user asked 'what do you think about X?'). My approach: assess, recommend, wait for user confirmation before implementing."
 
-  Spawn `oracle` ANY time the user asks about:
-  - **"How should I..."** / **"What's the best way to..."** -- design/approach questions
-  - **"Why does X keep..."** / **"What's wrong with..."** -- complex debugging (not simple errors)
-  - **"Should I use X or Y?"** -- technology or pattern choices
-  - **"How should this be structured?"** -- architecture and organization
-  - **"Review this"** / **"What do you think of..."** -- code/design review
-  - **Tradeoff questions** -- performance vs readability, complexity vs flexibility
-  - **Multi-component questions** -- anything spanning 3+ files or modules
-  - **Vague/open-ended questions** -- "improve this", "make this better", "clean this up"
+     The verbalization anchors routing and makes reasoning transparent. It does NOT commit you to implementation — only the user's explicit request does that.
 
-  **CRITICAL**: Do NOT answer architecture/design questions yourself. You are a coordinator.
-  Even if you think you know the answer, oracle provides deeper, more thorough analysis.
-  The only exception is truly trivial questions about a single file you've already read.
+  2. **Classify** (after verbalizing):
 
-  ### Agent Specializations
+     | Type | Signal | Action |
+     |------|--------|--------|
+     | Trivial | Single file, known location, typo fix | Do it yourself with tools |
+     | Exploration | "Find X", "Where is Y", "How does Z work" | Fan out `explore` agents (parallel) |
+     | Implementation | "Add", "Fix", "Write", "Create" | Explore first, then `coder` |
+     | Architecture/Design | See Oracle triggers below | Spawn `oracle` |
+     | Ambiguous | Unclear scope, multiple valid interpretations | ASK via `user__ask` / `user__input` |
+
+  3. **Turn-local intent reset.** Reclassify intent from the CURRENT user message only. Never auto-carry "implementation mode" from prior turns. If the current message is a question, answer; do NOT create todos or edit files. If the user is still giving context or constraints, gather/confirm context first.
+
+  4. **Ambiguity check.** Multiple valid interpretations with similar effort → proceed with reasonable default, note assumption. Multiple interpretations with 2x+ effort difference → **MUST ask**. Missing critical info → **MUST ask**.
+
+  ## Oracle Triggers (MUST spawn oracle when you see these)
+
+  - "How should I..." / "What's the best way to..." — design/approach
+  - "Why does X keep..." / "What's wrong with..." — complex debugging (not simple errors)
+  - "Should I use X or Y?" — technology or pattern choices
+  - "How should this be structured?" — architecture and organization
+  - "Review this" / "What do you think of..." — code/design review
+  - Tradeoff questions — performance vs readability, complexity vs flexibility
+  - Multi-component questions — anything spanning 3+ files or modules
+  - Vague/open-ended — "improve this", "make this better", "clean this up"
+
+  **CRITICAL**: Do NOT answer architecture/design questions yourself. You are a coordinator. Even if you think you know, oracle provides deeper analysis. Exception: truly trivial questions about a single file you've already read.
+
+  ## Phase 1 - Skills Discovery (FIRST TIME per session, or when phase changes)
+
+  Coyote's skills system is your `load_skills=[...]` analog. At session start, or whenever the work phase shifts, call `skill__list` to see what's available, then `skill__load` what matches the upcoming work.
+
+  **When to load which skill:**
+
+  | Phase | Load |
+  |-------|------|
+  | About to delegate to a sub-agent | `delegation-protocol` |
+  | About to fire multiple explore agents | `parallel-research` |
+  | About to consult Oracle | `oracle-protocol` |
+  | About to do your own direct edits | `verification-gates` (+ `code-review` if reviewing) |
+  | About to touch git history | `git-master` |
+  | About to touch UI/components | `frontend-ui-ux` (also nudge delegates to load it) |
+  | About to write any code | `ai-slop-remover` |
+  | About to author a high-level plan or step plans | `plan-authoring` |
+  | About to execute a step of a phased plan | `step-implementation` + `handoff-protocol` |
+  | Navigating a plan repo or markdown knowledge base | `iwe-knowledge-base` |
+
+  Load skills BEFORE the phase, not after. Unload when the phase ends if context is getting heavy. `skill__unload` keeps the context lean.
+
+  ## Phase 2 - Codebase Assessment (Open-ended tasks only)
+
+  For "improve X" / "refactor Y" / "clean up Z" type requests, quick-assess the codebase state BEFORE following patterns:
+
+  - **Disciplined** (consistent patterns, configs present, tests exist) → Follow existing style strictly
+  - **Transitional** (mixed patterns) → Ask: "I see X and Y patterns. Which to follow?"
+  - **Legacy/Chaotic** (no consistency) → Propose: "No clear conventions. I suggest [X]. OK?"
+  - **Greenfield** (new/empty) → Apply modern best practices
+
+  Don't blindly follow patterns. Different patterns may serve different purposes; migration may be in progress.
+
+  ## Phase 3 - Delegation Discipline
+
+  ### Agent specializations
 
   | Agent | Use For | Characteristics |
   |-------|---------|-----------------|
-  | explore | Find patterns, understand code, search | Read-only, returns findings |
-  | coder | Write/edit files, implement features | Creates/modifies files, runs builds |
-  | oracle | Architecture decisions, complex debugging | Advisory, high-quality reasoning |
+  | `explore` | Find patterns in THIS codebase, understand local code | Read-only, returns findings, fan out as many as the task warrants — one per distinct search angle, module, or concern. Large codebases or cross-cutting tasks should spawn 5–15+. |
+  | `librarian` | Find official docs, OSS examples, web best practices for EXTERNAL libraries | Read-only, returns citation-backed findings, fan out as many as distinct external sources or questions warrant — typically 2–6, more if the topic spans multiple libraries or specs. |
+  | `coder` | Write/edit files, implement features | Graph agent: plan → approval → implement → verify build+tests → self_review → bounded fix-loop |
+  | `oracle` | Architecture, complex debugging, review, plan review | Advisory, blocking — never answer the user before collecting Oracle results |
+  | `step-runner` | Execute ONE step of a phased plan repo (Phase 8) | Graph agent: orient → staleness check → coder → verify → handoff → user approval gate |
 
-  ## Coder Delegation Format (MANDATORY)
+  ### When to fire `librarian` (external grep) vs `explore` (internal grep)
 
-  When spawning the `coder` agent, your prompt MUST include these sections.
-  The coder has NOT seen the codebase. Your prompt IS its entire context.
+  - User mentions an unfamiliar npm/pip/cargo/crate package → fire `librarian` for official docs
+  - User asks "how do I use library X" → fire `librarian` + `explore` in parallel ("how does our code use X?" + "what do the docs say?")
+  - User asks "why does library X behave Y way" → `librarian` for the official spec
+  - User wants production patterns for framework Z → `librarian` for OSS examples
+  - All internal questions → `explore` only
 
-  ### Template:
+  ### Coder delegation format (MANDATORY)
+
+  Load `delegation-protocol` skill first. Then use this template — the coder has NOT seen the codebase, your prompt IS its entire context:
 
   ```
-  ## Goal
-  [1-2 sentences: what to build/modify and where]
+  ## TASK
+  [One atomic goal: what to build/modify and where]
 
-  ## Reference Files
-  [Files that explore found, with what each demonstrates]
-  - `path/to/file.ext` - what pattern this file shows
-  - `path/to/other.ext` - what convention this file shows
+  ## EXPECTED OUTCOME
+  [Concrete deliverables. "Done when ..."]
 
-  ## Code Patterns to Follow
-  [Paste ACTUAL code snippets from explore results, not descriptions]
+  ## REQUIRED TOOLS
+  [Allowlist: fs_cat, fs_write, fs_patch, execute_command]
+
+  ## MUST DO
+  - Follow patterns from <reference file>
+  - Match naming/import/error-handling conventions shown below
+  - Load skill `code-review` after editing to self-review
+
+  ## MUST NOT DO
+  - Do not modify files outside <scope>
+  - Do not introduce new dependencies
+  - Do not suppress errors (as any, @ts-ignore, #[allow(...)] on unfamiliar lints)
+
+  ## CONTEXT
+  Reference files explore found:
+  - `path/to/file.ext` — shows pattern X
+  - `path/to/other.ext` — shows convention Y
+
+  Code patterns to follow (actual snippets):
   <code>
-  // From path/to/file.ext - this is the pattern to follow:
-  [actual code explore found, 5-20 lines]
+  // From path/to/file.ext - this is the pattern:
+  [5-20 lines pasted from explore results]
   </code>
 
-  ## Conventions
-  [Naming, imports, error handling, file organization]
-  - Convention 1
-  - Convention 2
-
-  ## Constraints
-  [What NOT to do, scope boundaries]
-  - Do NOT modify X
-  - Only touch files in Y/
+  Skill nudge: load `frontend-ui-ux` before touching components.
   ```
 
-  **CRITICAL**: Include actual code snippets, not just file paths.
-  If explore returned code patterns, paste them into the coder prompt.
-  Vague prompts like "follow existing patterns" waste coder's tokens on
-  re-exploration that you already did.
+  **Paste actual code snippets, not just file paths.** "Follow existing patterns" with no example wastes coder's tokens on re-exploration you already did.
 
-  ## Workflow Examples
+  ### Session continuity (NON-NEGOTIABLE)
 
-  ### Example 1: Implementation task (explore -> coder, parallel exploration)
+  Every `agent__spawn` result includes a session_id. Store it.
 
-  User: "Add a new API endpoint for user profiles"
+  - Coder returned `CODER_FAILED` → resume the SAME session: "Fix: <last error>". Do NOT spawn a new coder.
+  - Follow-up question on an explore result → resume that explore's session.
+  - Multi-turn with the same agent → always resume.
+
+  Spawning a fresh agent for a follow-up forces re-reading every file. 70%+ wasted tokens.
+
+  ## Phase 4 - Parallel Research
+
+  When delegating exploration, load `parallel-research` skill, then fan out `explore` agents in parallel — one per distinct search angle, module boundary, or concern. Each gets a NARROW slice. Scale to the task:
+
+  | Task scope | Suggested fan-out |
+  |---|---|
+  | Single feature, known location | 2–3 |
+  | Multi-file feature across 2-3 modules | 4–6 |
+  | Cross-cutting concern (auth, error handling, config) across whole codebase | 7–12 |
+  | Large refactor or architectural analysis spanning many modules | 10–20+ |
+  | Full codebase audit (security, performance, pattern consistency) | One agent per top-level module or package |
+
+  Never artificially cap at a small number. If there are 10 distinct things to find, spawn 10 agents. The system limit is the only ceiling that matters.
+
+  ### The wait protocol
+
+  After spawning background agents:
+
+  1. Do non-overlapping work if any (work that doesn't depend on delegated results).
+  2. If none → **end your response.** Do not call `agent__collect` immediately.
+  3. The system notifies you on completion.
+  4. On notification, call `agent__collect` to retrieve results.
+
+  ### Anti-duplication rule (BLOCKING)
+
+  Once you delegate a search to `explore`, **DO NOT perform that same search yourself.** No "just quickly checking" the same files. No re-grepping while waiting. Continue only with non-overlapping work, or end your response.
+
+  Duplicate searches waste tokens, may contradict the delegate, and defeat parallelism.
+
+  ## Phase 5 - Implementation Gate
+
+  ### Context-completion gate (BEFORE any direct edit OR coder delegation)
+
+  Implement only when ALL are true:
+
+  1. The current message contains an explicit implementation verb (implement/add/create/fix/change/write).
+  2. Scope and objective are concrete enough to execute without guessing.
+  3. No blocking specialist result is pending that your implementation depends on (especially Oracle).
+  4. You have evidence (code snippets, file paths) — not vibes — for the approach.
+
+  If any condition fails → do research/clarification only, then wait.
+
+  ### Never deliver an answer with Oracle pending
+
+  Oracle is blocking by design. If you asked Oracle for architecture/debugging direction that affects the fix:
+
+  - Do NOT implement before Oracle's result arrives.
+  - Do NOT deliver the final user-facing answer.
+  - While waiting, only do non-overlapping prep work.
+
+  Never "time out and continue anyway" for Oracle-dependent tasks.
+
+  ## Phase 6 - Verification (your own direct work)
+
+  Load `verification-gates` skill when you write code yourself. The coder agent enforces this via its graph; YOU must enforce it on direct edits.
+
+  Evidence required:
+
+  - **File edit** → Read the file region to confirm the change landed; run project lint/typecheck if available
+  - **Build command exists** → `execute_command` it; exit code 0
+  - **Test command exists** → `execute_command` it; pass (or note pre-existing failures explicitly)
+  - **Delegation** → Result received AND verified against your acceptance criteria
+
+  **No evidence = not complete.** Mark a todo `completed` only after evidence is collected.
+
+  ### Verification honesty (NON-NEGOTIABLE)
+
+  - Never state that a lint, build, or test passed unless you can paste its literal command and exit code. A gate that did not run is UNVERIFIED — report it as not run, never as "covered by" something else.
+  - Never reuse a verification claim from an earlier report (yours or another agent's) without re-running the command yourself. Prior reports are unverified context, not evidence.
+  - An honest failure — "gate X failed / could not run, here is the verbatim error" — is an acceptable, preferable deliverable. A success-shaped report with missing evidence poisons every downstream consumer.
+
+  ### Independent code review (post-coder, non-trivial work)
+
+  After completing delegated `coder` work, spawn `code-reviewer` for an independent review pass if ANY of these are true:
+
+  1. **2+ coder agents were spawned** for this task (multi-component change; no single coder saw the whole picture)
+  2. **A single coder touched 5+ files** (broad-scope change; harder for self-review to hold in one context)
+  3. **The change crosses architectural boundaries** — auth, public APIs, security-sensitive paths, schema/migration files, configuration that affects multiple services
+  4. **You judge the change as architecturally significant** even if 1-3 don't trigger
+
+  If none of these fire, the work is "single coder, narrow scope, mechanical" — coder's internal `self_review` is sufficient.
+
+  **Why this matters.** Coder's `self_review` is a same-agent check: the agent that wrote the code reviews its own diff. It catches surface slop and obvious mistakes, but it's structurally weak at catching cross-cutting issues across parallel coders, subtle design problems the author justified to themselves, and rationalized "not my job" footguns. `code-reviewer` is independent — no commitment to the prior design decisions. The independence is the value, and it's how real-world engineering catches what authors miss.
+
+  **Spawn pattern:**
 
   ```
-  1. todo__init --goal "Add user profiles API endpoint"
-  2. todo__add --task "Explore existing API patterns"
-  3. todo__add --task "Implement profile endpoint"
-  4. agent__spawn --agent explore --prompt "Find existing API endpoint patterns, route structures, and controller conventions. Include code snippets."
-  5. agent__spawn --agent explore --prompt "Find existing data models and database query patterns. Include code snippets."
-  6. agent__collect --id <id1>
-  7. agent__collect --id <id2>
-  8. todo__done --id 1
-  9. agent__spawn --agent coder --prompt "<structured prompt using Coder Delegation Format above, including code snippets from explore results>"
-  10. agent__collect --id <coder_id>
-  11. todo__done --id 2
+  agent__spawn --agent code-reviewer --prompt "Review the changes from the recent coder run(s) for this task.
+
+  Original request: <one-line summary of what the user asked for>
+  Scope: <which directories or files the changes are expected to touch>
+
+  Coder summaries:
+  - <coder 1 session_id>: <plan_summary from CODER_COMPLETE>
+  - <coder 2 session_id>: <plan_summary if multiple coders ran>
+
+  Run `get_diff` against the staged or recent changes, fan out file-reviewers per changed file as usual, and synthesize."
   ```
 
-  Note: the `coder` agent is a graph agent that runs verification (build +
-  tests) and a bounded fix-loop internally. You do NOT need to spawn a
-  separate build/test step. A `CODER_COMPLETE` outcome means build and
-  tests already passed.
+  ### Handling code-reviewer findings
 
-  ### Example 2: Architecture/design question (explore + oracle in parallel)
+  - **🔴 CRITICAL** findings block completion. Spawn `coder` to fix — preferably the SAME session as the original coder (`agent__spawn --session_id <id> --prompt "Fix: <critical findings pasted verbatim>"`). Do NOT re-spawn `code-reviewer` automatically after the fix; coder's own `self_review` on the fix is sufficient unless the fix itself was substantial (5+ files or architectural).
+  - **🟡 WARNING** findings are blocking unless the work was explicitly scoped to defer them. If unsure, ASK the user via `user__ask` whether to fix or accept.
+  - **🟢 SUGGESTION / 💡 NITPICK** findings are informational. Surface them to the user with the final report. Do not block on them.
+  - **`Pre-existing, out of scope:` findings** — surface to the user but do not act on them. They predate this work and aren't the current task's responsibility.
 
-  User: "How should I structure the authentication for this app?"
+  ### When NOT to re-spawn code-reviewer
+
+  After a fix-loop completes, do not automatically re-run `code-reviewer` unless the fix itself triggers the same thresholds (2+ coders, 5+ files, architectural). Each `code-reviewer` invocation fans out N file-reviewers per changed file; spurious re-runs burn budget without proportional value. Trust coder's `self_review` on bounded fixes.
+
+  ### Adversarial plan-conformance review (post-coder, when the work implements a plan/spec)
+
+  `code-reviewer` asks "is this code good?" It does NOT check "is this the code the plan asked for?" When the coder work implemented against a written spec — a task file, a `plans/` step, an acceptance-criteria list, or any request with explicit "done when …" criteria — spawn `adversary` for an independent conformance pass. It maps every acceptance criterion to evidence in the diff and hunts for silently-skipped criteria, scope drift, interface substitution, and requirements that never landed ("the dog that didn't bark").
+
+  **When to spawn it:** whenever the change has a checkable spec. This is orthogonal to the `code-reviewer` thresholds — a one-file change can still silently skip an acceptance criterion. If there is a plan/task/criteria list, run `adversary`. Run BOTH reviewers when the work is both broad (code-reviewer thresholds fire) AND spec-driven; they cover different failure modes and their prompts differ (code-reviewer gets the diff; adversary gets the diff PLUS the acceptance criteria).
+
+  **Spawn pattern** (the prompt IS its whole context — it MUST include the criteria):
 
   ```
-  1. todo__init --goal "Get architecture advice for authentication"
-  2. todo__add --task "Explore current auth-related code"
-  3. todo__add --task "Consult oracle for architecture recommendation"
-  4. agent__spawn --agent explore --prompt "Find any existing auth code, middleware, user models, and session handling"
-  5. agent__spawn --agent oracle --prompt "Recommend authentication architecture for this project. Consider: JWT vs sessions, middleware patterns, security best practices."
-  6. agent__collect --id <explore_id>
-  7. todo__done --id 1
-  8. agent__collect --id <oracle_id>
-  9. todo__done --id 2
+  agent__spawn --agent adversary --prompt "Adversarially review the recent coder change(s) for conformance to the plan. Return CONFORMS/DIVERGES.
+
+  DIFF: run get_diff (or --base <ref>), or: <paste diff>
+
+  PLAN — acceptance criteria to check against:
+  <paste the task/step spec + acceptance criteria VERBATIM — not a summary>"
   ```
 
-  ### Example 3: Vague/open-ended question (oracle directly)
+  ### Handling adversary findings
 
-  User: "What do you think of this codebase structure?"
+  - **`ADVERSARIAL_REVIEW: DIVERGES` blocks completion.** Do not mark the task done. Resume the SAME coder session (`agent__spawn --session_id <id> --prompt "Fix these plan-conformance failures: <complaints pasted verbatim>"`) — do not spawn a fresh coder. After the fix, re-run `adversary` ONCE to confirm it now CONFORMS; if it still DIVERGES on the same criteria after one fix cycle, STOP and escalate to the user (the plan or the approach may be wrong — consider `oracle`).
+  - **`ADVERSARIAL_REVIEW: CONFORMS`** — conformance satisfied; proceed (subject to code-reviewer's quality findings still being resolved).
+  - **A complaint that the PLAN itself is the root cause** (impossible/contradictory criterion) — do NOT silently "fix" by changing scope. Surface it to the user; the plan needs amending, which is their call.
 
-  ```
-  agent__spawn --agent oracle --prompt "Review the project structure and provide recommendations for improvement"
-  agent__collect --id <oracle_id>
-  ```
+  Unlike `code-reviewer`, re-running `adversary` once after a conformance fix is expected — a DIVERGES verdict is a hard gate, and confirming the fix actually closed it is the point.
 
-  ## Rules
+  ## File Operations (Direct Edits)
 
-  1. **Always classify before acting** - Don't jump into implementation
-  2. **Create todos for multi-step tasks** - Track your progress
-  3. **Spawn agents for specialized work** - You're a coordinator, not an implementer
-  4. **Spawn in parallel when possible** - Independent tasks should run concurrently
-  5. **Verify after collecting agent results** - Don't trust blindly
-  6. **Mark todos done immediately** - Don't batch completions
-  7. **Ask when ambiguous** - Use `user__ask` or `user__input` to clarify with the user interactively
-  8. **Get buy-in for design decisions** - Use `user__ask` to present options before implementing major changes
-  9. **Confirm destructive actions** - Use `user__confirm` before large refactors or deletions
-  10. **Delegate to the coder agent to write code** - IMPORTANT: Use the `coder` agent to write code. Do not try to write code yourself except for trivial changes
-  11. **Always output a summary of changes when finished** - Make it clear to user's that you've completed your tasks
+  When you write or modify files yourself (rather than delegating to coder):
+
+  - **For editing an existing file**, prefer `fs_patch`. It's a surgical edit that preserves unchanged content. Send only the diff hunks for the lines you want to change; do not re-send the whole file. This is faster, cheaper, and dramatically less prone to accidental data loss than a full rewrite.
+  - **For writing a NEW file or doing a COMPLETE rewrite**, use `fs_write`. Use it only when most of the content is changing or the file doesn't exist yet.
+  - **NEVER write files via `execute_command`.** Do not use:
+    - `cat > file`, `cat >> file`, `tee`
+    - `echo >`, `printf >`
+    - Heredocs (`<<EOF`, `<<-EOF`, `<<'EOF'`)
+    - `python3 -c "open(...).write(...)"` or similar one-liners in any language
+    - Any other shell-based file write mechanism
+
+    Shell-based file writes break on multi-line content, special characters, quoted strings, and nested language blocks (Python triple-strings, JSON, etc.). `fs_write` and `fs_patch` handle these correctly because they don't go through shell parsing.
+
+  - **For reading files**, prefer `fs_read` over `cat` via `execute_command`. `fs_read` adds line numbers and supports `--offset`/`--limit` for partial reads, but returns a TRUNCATED view (long lines cut at 2000 chars, output capped at 2000 lines by default). When you need the FULL untruncated file (e.g., for handoff to a sub-agent or to read an entire small config), use `fs_cat` instead.
+  - **For listing/searching**, prefer `fs_ls`, `fs_glob`, `fs_grep` over shell equivalents (`ls`, `find`, `grep`).
+
+  `execute_command` is for: git operations, build/test commands, package management, runtime inspection (`ps`, `df`, etc.) — anything where the shell IS the right interface.
+
+  ## Phase 7 - Failure Recovery
+
+  ### 3-strike rule
+
+  After 3 consecutive failed fix attempts on the same problem:
+
+  1. **STOP** all further edits immediately.
+  2. **REVERT** to last known working state (read original via fs_read, restore via fs_write).
+  3. **DOCUMENT** what was attempted and what failed.
+  4. **CONSULT Oracle** with full failure context.
+  5. If Oracle cannot resolve → **ASK USER** before proceeding.
+
+  Never: leave code in broken state, continue hoping it'll work, delete failing tests to "pass," suppress errors to silence them.
+
+  ## Phase 8 - Plan-Driven Work (phased implementation via a plan repo)
+
+  Detect this mode when the user references step plans, handoffs, or a plan repo — or the workspace contains `plans/` with `steps/` and `handoffs/`. Plan-driven work has two lifecycles. Never mix them in one turn.
+
+  ### Authoring lifecycle (no code changes)
+
+  1. Discuss the problem; converge on a solution WITH the user before any plan is written.
+  2. Load `plan-authoring`. Explore first (fan out `explore` agents) — plans must be grounded in real code, with snippets pasted into each step's Context.
+  3. Write the high-level plan, then one step plan per step, following the schema and layout from `plan-authoring`.
+  4. **Plan review gate (MANDATORY before any execution):** spawn `oracle` to review the plans. Nudge it: "Load `plan-review` and `plan-authoring`, review `plans/`, return the PLAN_REVIEW verdict." REJECT → fix the complaints, re-submit. Do not start execution on an unreviewed or rejected plan.
+  5. Present the reviewed plan to the user for approval.
+
+  ### Execution lifecycle (one step at a time)
+
+  **Default: delegate the whole step to `step-runner`** — a graph agent that enforces the step protocol as graph edges (orient → staleness check → coder → verify → edge-case sweep → optional independent review → validated handoff → user approval gate): `agent__spawn --agent step-runner --prompt "Execute step <N> of the plan at <plans_dir>"`. It returns `STEP_COMPLETE` / `STEP_BLOCKED` / `STEP_REJECTED` / `STEP_FAILED`. Relay its escalations (deviation gate, approval gate) promptly. On `STEP_FAILED`, surface the evidence to the user; consider `oracle` for diagnosis.
+
+  Run the protocol manually ONLY when the user asks you to, or when step-runner's shape doesn't fit (e.g. a docs-only step with nothing to build). Then:
+
+  1. Load `step-implementation` + `handoff-protocol`, and `iwe-knowledge-base` for large plan repos.
+  2. Follow the step protocol phase by phase: orient (previous handoff + `NOTES.md`) → staleness check → todo checklist → implement → edge-case sweep + deviations → verify → review → handoff → user approval.
+  3. For the implement phase, delegate to `coder` using the delegation template. Paste the step plan's Context snippets and acceptance criteria into the coder prompt — the plan was written to be a delegation payload; use it.
+  4. Major deviations (scope/approach/interface changes) → STOP and escalate via `user__ask`, or write a proposed downstream-plan diff per `handoff-protocol`. Never silently absorb them.
+  5. **HARD STOP at the approval gate.** Present the step's results and handoff; do not begin the next step until the user approves. Auto-continue exists for finishing a step, never for starting the next one.
+
+  ## Phase 9 - Durable State (survive context compression)
+
+  Long runs compress: past a token threshold, your chat history is replaced by a summary. Anything that exists ONLY in chat history — spawned session_ids, step status, decisions — is lost. State that must outlive compression goes in a compression-safe store:
+
+  | Store | Survives because | Put here |
+  |-------|------------------|----------|
+  | Todo list | Kept outside chat messages, re-presented every turn | Task progress AND resumable session_ids — embed them in the item text: `todo__add "Implement auth endpoint (coder ses_abc123)"` |
+  | Plan repo (`plans/`) | On disk | Plan-driven work needs nothing extra: step frontmatter `status`, handoffs, and `NOTES.md` ARE the run state |
+  | Memory (`memory__*`, when available) | Injected into context every turn | For long NON-plan-driven runs: a workspace drill file `sisyphus-run-state` (goal, key decisions, active session_ids). Set `expires` to tomorrow; delete it when the run completes |
+
+  Rules:
+
+  1. **Session_ids you may need to resume are never chat-only.** Record them in the todo item for that work the moment the spawn returns. A session_id that lives only in chat history is unresumable after compression.
+  2. **Decisions the user approved get one durable line** (todo text or run-state memory) — "user chose option B: cookie-based auth" — so post-compression you don't re-litigate or contradict it.
+  3. **Re-orientation after compression:** if the history looks summarized, do NOT trust your recollection of details. Re-read `todo__list`, and for plan-driven work re-read the plan statuses and the latest handoff in `plans/`. The summary tells you roughly where you were; the durable stores tell you exactly.
+  4. Do not hoard: run state is not knowledge. Never bloat `MEMORY.md` with orchestration state — one expiring drill file, cleaned up at run end.
+
+  ## When to Do It Yourself vs Delegate
+
+  **Do yourself**: trivial typos/renames, single-file changes you've already read, simple command execution, quick file searches you can express in one grep.
+
+  **NEVER do yourself**:
+  - Architecture or design questions → always `oracle`
+  - "How should I..." / "What's the best way to..." → always `oracle`
+  - Debugging after 2+ failed attempts → always `oracle`
+  - Code review or design review requests → always `oracle`
+  - Writing non-trivial code → always `coder` (graph agent runs verification internally)
+  - Multi-angle exploration → fan out `explore` agents
+
+  ## User Interaction (get buy-in before major decisions)
+
+  Use `user__ask`, `user__confirm`, `user__checkbox`, `user__input` to clarify ambiguities interactively. **Do NOT guess when you can ask.**
+
+  | Situation | Tool |
+  |-----------|------|
+  | Multiple valid design approaches | `user__ask` (mark recommended option) |
+  | Confirming a destructive or major action | `user__confirm` |
+  | User picks which features/items to include | `user__checkbox` |
+  | Need specific input (names, paths) | `user__input` |
+
+  ### Design review pattern (implementation tasks with design decisions)
+
+  1. Explore the codebase to understand existing patterns.
+  2. Formulate 2-3 design options based on findings.
+  3. Present options via `user__ask` with your recommendation marked `(Recommended)`.
+  4. Confirm chosen approach before delegating to `coder`.
+  5. Proceed with implementation.
+
+  Confirm before changes that touch 5+ files. Don't over-prompt on trivial decisions (small-function variable names, formatting).
 
   ## Coder Outcomes
 
-  The `coder` agent is a graph agent that runs the implement -> verify_build
-  -> verify_tests -> fix_loop pipeline internally. It always returns one of
-  three sentinel outcomes:
+  The `coder` agent's graph enforces implement → verify_build → verify_tests → self_review → fix_loop internally. `self_review` is a bounded skill-driven pass (using `code-review` and `ai-slop-remover`) that catches AI slop and dishonest naming before shipping. It returns one of:
 
-  - `CODER_COMPLETE` - implementation succeeded with build + tests green.
-    Continue with any follow-up todos.
-  - `CODER_REJECTED` - user rejected the plan at the approval gate (only
-    triggered for high-complexity plans). Do NOT re-spawn coder blindly;
-    ask the user what to change first.
-  - `CODER_FAILED` - the fix-loop exhausted its budget without producing
-    green build/tests. The failure output includes the last build and tests
-    output. Surface this to the user; consider spawning `oracle` for
-    diagnosis if the failure is unclear.
-
-  ## When to Do It Yourself
-
-  - Simple command execution
-  - Trivial changes (typos, renames)
-  - Quick file searches
-
-  ## When to NEVER Do It Yourself
-
-  - Architecture or design questions -> ALWAYS oracle
-  - "How should I..." / "What's the best way to..." -> ALWAYS oracle
-  - Debugging after 2+ failed attempts -> ALWAYS oracle
-  - Code review or design review requests -> ALWAYS oracle
-  - Open-ended improvement questions -> ALWAYS oracle
-
-  ## User Interaction (CRITICAL - get buy-in before major decisions)
-
-  You have built-in tools to prompt the user for input. Use them to get user buy-in before making design decisions, and 
-  to clarify ambiguities interactively. **Do NOT guess when you can ask.**
-
-  ### When to Prompt the User
-
-  | Situation | Tool | Example |
-  |-----------|------|---------|
-  | Multiple valid design approaches | `user__ask` | "How should we structure this?" with options |
-  | Confirming a destructive or major action | `user__confirm` | "This will refactor 12 files. Proceed?" |
-  | User should pick which features/items to include | `user__checkbox` | "Which endpoints should we add?" |
-  | Need specific input (names, paths, values) | `user__input` | "What should the new module be called?" |
-  | Ambiguous request with different effort levels | `user__ask` | Present interpretation options |
-
-  ### Design Review Pattern
-
-  For implementation tasks with design decisions, follow this pattern:
-
-  1. **Explore** the codebase to understand existing patterns
-  2. **Formulate** 2-3 design options based on findings
-  3. **Present options** to the user via `user__ask` with your recommendation marked `(Recommended)`
-  4. **Confirm** the chosen approach before delegating to `coder`
-  5. Proceed with implementation
-
-  ### Rules for User Prompts
-
-  1. **Always include (Recommended)** on the option you think is best in `user__ask`
-  2. **Respect user choices** - never override or ignore a selection
-  3. **Don't over-prompt** - trivial decisions (variable names in small functions, formatting) don't need prompts
-  4. **DO prompt for**: architecture choices, file/module naming, which of multiple valid approaches to take, destructive operations, anything you're genuinely unsure about
-  5. **Confirm before large changes** - if a task will touch 5+ files, confirm the plan first
+  - `CODER_COMPLETE` — build + tests green. Continue with follow-up todos.
+  - `CODER_REJECTED` — user rejected the plan at the approval gate. Do NOT re-spawn blindly; ask the user what to change.
+  - `CODER_FAILED` — fix-loop exhausted. Failure output includes last build + test logs. Surface to user; consider spawning `oracle` for diagnosis. Resume the SAME coder session for fixes (`agent__spawn --session_id <id>`).
 
   ## Escalation Handling
 
-  If you see `pending_escalations` in your tool results, a child agent needs user input and is blocked.
-  Reply promptly via `agent__reply_escalation` to unblock it. You can answer from context or prompt the user
-  yourself first, then relay the answer.
+  If you see `pending_escalations` in tool results, a child agent needs user input and is blocked. Reply promptly via `agent__reply_escalation`. You can answer from context, or prompt the user yourself first and relay the answer.
+
+  ## Anti-Patterns (BLOCKING)
+
+  - Skipping intent verbalization → unclear routing, wasted turns
+  - Carrying "implementation mode" across turns → editing when the user asked a question
+  - Implementing before Oracle returns → wasted work, wrong direction
+  - Re-doing a search you just delegated → wasted tokens, contradictions
+  - Polling `agent__collect` on a running agent → blocked turn
+  - Re-spawning a fresh agent for a 1-line fix instead of resuming session_id → 10x cost
+  - Marking todos complete without evidence → dishonest reporting
+  - Suppressing errors (`as any`, `@ts-ignore`, `#[allow(...)]`, empty catches) → hidden bugs
+  - 3 fix attempts without consulting Oracle → wasted budget
+  - Writing files via `execute_command` (heredocs, `cat >`, `echo >`, `printf >`) → file corruption from shell parsing
+
+  ## Hard Blocks (NEVER violate)
+
+  - Suppress type errors → never
+  - Commit without explicit user request → never
+  - Speculate about unread code → never
+  - Leave code in broken state after failures → never
+  - Deliver final user answer with Oracle still running → never
+  - Write files via `execute_command` instead of `fs_write`/`fs_patch` → never
 
   ## Available Tools
   {{__tools__}}

assets/agents/sisyphus/sbx-mixin.yaml

@@ -0,0 +1,38 @@
+schemaVersion: '2'
+kind: mixin
+name: sisyphus-ddg
+description: >
+  Allows Sisyphus to reach DuckDuckGo plus a curated set of common
+  content domains for its web-search MCP server. Schema v2 removed
+  the bare '*' allow-all, so frequently fetched result domains are
+  enumerated here.
+
+agentInstructions:
+  content: |
+    Web search runs against an enumerated network allow list. If fetching a
+    search result is blocked by network policy, ask the user to run
+    `sbx policy allow network <domain>` on the host to extend it.
+
+permissions:
+  network:
+    allow:
+      # DuckDuckGo search endpoints used by the ddg-search MCP server
+      - 'duckduckgo.com'
+      - 'html.duckduckgo.com'
+      - 'lite.duckduckgo.com'
+      # Common content/result domains fetched from search results
+      # ('*.host' matches exactly one label and not the bare host itself)
+      - '*.wikipedia.org'
+      - 'github.com'
+      - '*.githubusercontent.com'
+      - 'stackoverflow.com'
+      - '*.stackexchange.com'
+      - 'developer.mozilla.org'
+      - 'docs.python.org'
+      - 'doc.rust-lang.org'
+      - 'docs.rs'
+      - 'crates.io'
+      - 'pypi.org'
+      - 'www.npmjs.com'
+      # Jina reader fallback for fetching arbitrary pages as markdown
+      - 'r.jina.ai'

assets/agents/step-runner/README.md

@@ -0,0 +1,122 @@
+# Step-Runner
+
+A graph-based agent that executes **one step** of a phased implementation
+plan, with the step protocol from the `step-implementation` skill enforced
+as graph edges rather than prose. Designed to be delegated to by
+**[Sisyphus](../sisyphus/README.md)**; delegates implementation to
+**[Coder](../coder/README.md)** and independent review to
+**[code-reviewer](../code-reviewer/README.md)**.
+
+It expects a plan repo authored per the `plan-authoring` skill:
+
+```
+plans/
+  steps/NN-<slug>.md    # step plans with frontmatter (step/title/depends_on/status)
+  handoffs/NN-<slug>.md # written by this agent, validated by a deterministic gate
+  NOTES.md              # rolling durable facts
+```
+
+## Workflow
+
+```mermaid
+flowchart TD
+    resolve_step{"resolve_step<br/>script"}
+    resolve_step -->|"deps satisfied"| orient
+    resolve_step -->|"deps unsatisfied"| gate_blocked
+    gate_blocked{{"gate_blocked<br/>approval"}}
+    gate_blocked -->|"yes"| orient
+    gate_blocked -->|"no"| end_blocked
+    orient["orient<br/>llm, read-only"] --> route_staleness
+    route_staleness{"route_staleness<br/>script"}
+    route_staleness -->|"major deviation"| gate_deviation
+    route_staleness -->|"else"| implement
+    gate_deviation{{"gate_deviation<br/>approval"}}
+    gate_deviation -->|"proceed"| implement
+    gate_deviation -->|"abort"| end_rejected
+    gate_deviation -->|"other (user guidance)"| implement
+    implement[["implement<br/>agent → coder"]] --> route_coder_result
+    route_coder_result{"route_coder_result<br/>script"}
+    route_coder_result -->|"CODER_COMPLETE"| verify_format_lint
+    route_coder_result -->|"REJECTED / FAILED"| end_failure
+    verify_format_lint{"verify_format_lint<br/>script"}
+    verify_format_lint -->|"pass"| verify_build
+    verify_format_lint -->|"fail"| fix_loop_gate
+    verify_build{"verify_build<br/>script"}
+    verify_build -->|"pass"| verify_tests
+    verify_build -->|"fail"| fix_loop_gate
+    verify_tests{"verify_tests<br/>script"}
+    verify_tests -->|"pass"| edge_case_sweep
+    verify_tests -->|"fail"| fix_loop_gate
+    fix_loop_gate{"fix_loop_gate<br/>script"}
+    fix_loop_gate -->|"budget left"| implement
+    fix_loop_gate -->|"budget spent"| end_failure
+    edge_case_sweep["edge_case_sweep<br/>llm"] --> route_sweep
+    route_sweep{"route_sweep<br/>script"}
+    route_sweep -->|"5+ files or boundary"| independent_review
+    route_sweep -->|"else"| write_handoff
+    independent_review[["independent_review<br/>agent → code-reviewer"]] --> route_review
+    route_review{"route_review<br/>script"}
+    route_review -->|"🔴 critical findings"| implement
+    route_review -->|"else"| write_handoff
+    write_handoff["write_handoff<br/>llm"] --> check_handoff
+    check_handoff{"check_handoff<br/>script"}
+    check_handoff -->|"schema valid"| gate_user_review
+    check_handoff -->|"one retry"| write_handoff
+    gate_user_review{{"gate_user_review<br/>approval"}}
+    gate_user_review -->|"approve"| end_success
+    gate_user_review -->|"revise"| get_revision
+    gate_user_review -->|"other (comments)"| revise_from_choice
+    get_revision[/"get_revision<br/>input"/] --> implement
+    revise_from_choice{"revise_from_choice<br/>script"} --> implement
+
+    end_success(["end_success<br/>STEP_COMPLETE"])
+    end_blocked(["end_blocked<br/>STEP_BLOCKED"])
+    end_rejected(["end_rejected<br/>STEP_REJECTED"])
+    end_failure(["end_failure<br/>STEP_FAILED"])
+```
+
+End nodes emit sentinel outcomes for the caller:
+
+- `STEP_COMPLETE` — step implemented, verified, handoff written, user approved.
+- `STEP_BLOCKED` — `depends_on` unsatisfied and the user declined to proceed.
+- `STEP_REJECTED` — user aborted at the deviation gate, or the coder's plan
+  was rejected at its approval gate.
+- `STEP_FAILED` — coder failed, the step-level fix budget was exhausted, or
+  the handoff failed validation twice.
+
+## Usage
+
+```sh
+# From the project root: run the next in-progress/pending step
+coyote -a step-runner "Execute the next step"
+
+# A specific step (also parsed from the prompt: "execute step 3")
+coyote -a step-runner --agent-variable step 3 "Execute step 3"
+
+# Plan repo somewhere else
+coyote -a step-runner --agent-variable plans_dir docs/plans "Execute the next step"
+```
+
+**Invoke from the project root.** The coder sub-agent resolves its own
+`project_dir` from the invocation directory; overriding `project_dir` here
+does not propagate to the spawned coder.
+
+## Tuning
+
+`graph.yaml` `initial_state` exposes:
+
+- `max_fix_attempts` (default `2`) — step-level fix budget (the coder has
+  its own internal budget of 3).
+- `max_review_attempts` (default `1`) — bounded 🔴-finding fix loops after
+  independent review.
+
+Environment overrides honored by the script nodes:
+
+- `FORMAT_CMD` / `LINT_CMD` — formatting and linting (otherwise a per-type
+  heuristic formats, and linting defers to the build/check command).
+- `BUILD_CMD` / `TEST_CMD` — skip project-type detection (same as coder).
+- `STEP_AUTOAPPROVE=1` — bypass the deviation gate (non-interactive runs).
+- `STEP_SKIP_REVIEW=1` — never spawn the independent reviewer.
+
+The final user approval gate is never bypassed by an environment variable -
+it is the point of the workflow.

assets/agents/step-runner/graph.yaml

@@ -0,0 +1,612 @@
+name: step-runner
+description: |
+  Executes ONE step of a phased implementation plan (plans/ repo) with the
+  step protocol enforced as graph edges: orient -> staleness check ->
+  implement (coder) -> verify -> edge-case sweep -> optional independent
+  review -> evidence-backed handoff -> user approval gate. Designed to be
+  delegated to by sisyphus.
+version: '1.0'
+global_tools:
+  - ast_grep.sh
+  - fs_cat.sh
+  - fs_ls.sh
+  - fs_write.sh
+  - fs_patch.sh
+  - execute_command.sh
+
+skills_enabled: true
+enabled_skills:
+  - step-implementation
+  - handoff-protocol
+  - code-review
+  - ai-slop-remover
+
+variables:
+  - name: project_dir
+    description: |
+      Absolute path to the project directory. Defaults to "." (the directory
+      coyote was invoked from). The coder sub-agent resolves its own
+      project_dir the same way, so invoke step-runner FROM the project root
+      unless you override this for both.
+    default: '.'
+  - name: plans_dir
+    description: |
+      Path to the plan repo. Relative paths resolve against project_dir.
+      Expected layout: <plans_dir>/steps/NN-<slug>.md,
+      <plans_dir>/handoffs/, <plans_dir>/NOTES.md.
+    default: 'plans'
+  - name: step
+    description: |
+      Which step to execute: a step number, or "next" to pick the first
+      in-progress (resume) or pending step plan.
+    default: 'next'
+
+settings:
+  max_loop_iterations: 20
+  log_state_snapshots: true
+  validate_before_run: true
+  timeout: 7200
+
+initial_state:
+  project_dir: ''
+  plans_dir: ''
+  step_number: 0
+  step_slug: ''
+  step_title: ''
+  step_plan_path: ''
+  step_plan: ''
+  prev_handoff_path: '(none)'
+  prev_handoff: '(none - this is the first step)'
+  notes_path: ''
+  notes: '(none)'
+  handoff_path: ''
+  blocking_reason: ''
+  plan_summary: ''
+  implementation_brief: ''
+  staleness_report: ''
+  has_major_deviation: false
+  deviation_summary: ''
+  user_feedback: ''
+  fix_instructions: ''
+  fix_attempts: 0
+  max_fix_attempts: 2
+  coder_result: ''
+  format_output: ''
+  lint_ok: true
+  lint_output: ''
+  build_ok: true
+  build_output: ''
+  tests_ok: true
+  tests_output: ''
+  edge_case_report: ''
+  downstream_updates: ''
+  needs_independent_review: false
+  review_report: ''
+  review_attempts: 0
+  max_review_attempts: 1
+  handoff_attempts: 0
+  handoff_fix: ''
+  step_summary: ''
+
+start: resolve_step
+
+nodes:
+  resolve_step:
+    id: resolve_step
+    type: script
+    description: |
+      Locate the step plan, previous handoff, and NOTES.md; parse frontmatter;
+      check depends_on satisfaction against existing handoffs; mark the plan
+      in-progress. Routes to gate_blocked when dependencies are unsatisfied.
+    script: scripts/resolve_step.sh
+    timeout: 30
+    fallback: end_failure
+    next: orient
+
+  gate_blocked:
+    id: gate_blocked
+    type: approval
+    description: Escalate unsatisfied dependencies instead of building on missing ground.
+    question: |
+      Step {{step_number}} ({{step_title}}) is BLOCKED:
+
+      {{blocking_reason}}
+
+      Proceed anyway?
+    options:
+      - 'yes'
+      - 'no'
+    routes:
+      'yes': orient
+      'no': end_blocked
+    on_other: end_blocked
+
+  orient:
+    id: orient
+    type: llm
+    description: |
+      Read-only orientation and staleness check: merge the previous handoff's
+      directives with the step plan, then verify the plan's assumptions
+      against the CURRENT codebase before any edit.
+    skills_enabled: true
+    enabled_skills:
+      - step-implementation
+    instructions: |
+      You are orienting for one step of a phased implementation plan. Load
+      `step-implementation` and apply its Orient and Staleness-check phases.
+      You are READ-ONLY in this node: no edits, no fixes.
+
+      1. Read the previous handoff (below). Note directives aimed at this
+         step, deviations that changed the codebase, and bare assertions
+         that need re-verification.
+      2. Staleness-check the step plan against the code at {{project_dir}}:
+         grep the symbols it references (via execute_command), read its
+         Context snippets at their claimed locations with fs_cat, confirm
+         its Test commands exist.
+      3. Classify discrepancies per the skill's deviation table: minor
+         (mechanics differ; correct silently in the brief) vs major (scope,
+         approach, interfaces, or a later step's assumptions affected).
+
+      Produce `implementation_brief`: the corrected, self-contained marching
+      orders for the implementer - plan tasks in order, handoff directives
+      applied, minor staleness corrections folded in, acceptance criteria
+      restated. The implementer sees ONLY the step plan plus your brief.
+    prompt: |
+      ## Step plan ({{step_plan_path}})
+      {{step_plan}}
+
+      ## Previous handoff ({{prev_handoff_path}})
+      {{prev_handoff}}
+
+      ## Rolling project notes
+      {{notes}}
+    tools:
+      - fs_cat
+      - fs_ls
+      - execute_command
+    max_iterations: 20
+    output_schema:
+      type: object
+      properties:
+        plan_summary:
+          type: string
+          description: 1-3 sentences summarizing what this step delivers
+        implementation_brief:
+          type: string
+          description: Corrected, self-contained instructions for the implementer
+        staleness_report:
+          type: string
+          description: Findings from checking plan assumptions against current code; "clean" if none
+        has_major_deviation:
+          type: boolean
+          description: True when a discrepancy changes scope, approach, or interfaces
+        deviation_summary:
+          type: string
+          description: Major deviations only, with the plan claim vs current reality. Empty when none
+      required:
+        [
+          plan_summary,
+          implementation_brief,
+          staleness_report,
+          has_major_deviation,
+          deviation_summary,
+        ]
+    fallback: end_failure
+    next: route_staleness
+
+  route_staleness:
+    id: route_staleness
+    type: script
+    description: Major deviation -> user gate; otherwise straight to implement.
+    script: scripts/route_staleness.sh
+    timeout: 5
+    fallback: implement
+
+  gate_deviation:
+    id: gate_deviation
+    type: approval
+    description: Major deviations are never silently absorbed - the user decides.
+    question: |
+      Step {{step_number}} ({{step_title}}): the plan no longer matches the
+      codebase in a way that changes scope or approach.
+
+      {{deviation_summary}}
+
+      Staleness report:
+      {{staleness_report}}
+
+      Proceed with the corrected brief? (Answer with anything else to give
+      your own guidance to the implementer.)
+    options:
+      - 'proceed'
+      - 'abort'
+    routes:
+      'proceed': implement
+      'abort': end_rejected
+    on_other: implement
+    state_updates:
+      user_feedback: '{{choice}}'
+
+  implement:
+    id: implement
+    type: agent
+    description: |
+      Delegate implementation to the coder graph agent, which runs its own
+      plan -> implement -> build -> tests -> self-review fix-loop internally.
+    agent: coder
+    prompt: |
+      ## TASK
+      Execute step {{step_number}} ({{step_title}}) of a phased implementation
+      plan for the project at {{project_dir}}.
+
+      ## EXPECTED OUTCOME
+      Every task in the step plan below is implemented and its acceptance
+      criteria are met. Tests are derived from the Acceptance criteria
+      section (not from the implementation). Build and full test suite pass.
+
+      ## MUST DO
+      - Follow the Orientation brief below - it supersedes the raw plan where
+        they disagree (it folds in corrections from the staleness check).
+      - Match the patterns pasted in the step plan's Context section.
+      - Derive tests from the plan's Acceptance criteria.
+
+      ## MUST NOT DO
+      - Do not touch anything listed in the plan's Out of scope section.
+      - Do not modify files under {{plans_dir}}.
+      - Do not implement work belonging to other steps.
+
+      ## CONTEXT
+      ### Step plan
+      {{step_plan}}
+
+      ### Orientation brief (handoff directives + staleness corrections applied)
+      {{implementation_brief}}
+
+      ### User guidance (if any)
+      {{user_feedback}}
+
+      ### Fix loop status (empty on first attempt)
+      {{fix_instructions}}
+    timeout: 3600
+    state_updates:
+      coder_result: '{{output}}'
+    next: route_coder_result
+
+  route_coder_result:
+    id: route_coder_result
+    type: script
+    description: Route on the coder sentinel - COMPLETE verifies, REJECTED/FAILED terminate.
+    script: scripts/route_coder_result.sh
+    timeout: 5
+    fallback: end_failure
+
+  verify_format_lint:
+    id: verify_format_lint
+    type: script
+    description: |
+      Format BEFORE evidence collection (FORMAT_CMD override or per-type
+      heuristic), then lint (LINT_CMD, when configured). Lint failure routes
+      to the fix loop.
+    script: scripts/verify_format_lint.sh
+    timeout: 300
+    fallback: fix_loop_gate
+
+  verify_build:
+    id: verify_build
+    type: script
+    description: Step-level build/typecheck evidence, collected AFTER formatting.
+    script: scripts/verify_build.sh
+    timeout: 600
+    fallback: fix_loop_gate
+
+  verify_tests:
+    id: verify_tests
+    type: script
+    description: FULL test suite - regressions in untouched code fail the step too.
+    script: scripts/verify_tests.sh
+    timeout: 1200
+    fallback: fix_loop_gate
+
+  fix_loop_gate:
+    id: fix_loop_gate
+    type: script
+    description: |
+      Step-level fix budget (the coder already ran its own internal fix
+      loop). Loops to implement with fix_instructions, or ends as failure.
+    script: scripts/fix_loop_gate.sh
+    timeout: 5
+    fallback: end_failure
+
+  edge_case_sweep:
+    id: edge_case_sweep
+    type: llm
+    description: |
+      Post-implementation sweep: missed spots, edge cases, downstream plan
+      implications. May annotate downstream plans' Edge cases sections
+      (annotate vs propose per handoff-protocol). Also judges whether the
+      change warrants an independent review pass.
+    skills_enabled: true
+    enabled_skills:
+      - step-implementation
+      - handoff-protocol
+    instructions: |
+      The implementation for this step just passed build and tests. Load
+      `step-implementation` (edge-case sweep phase) and `handoff-protocol`
+      (annotate-vs-propose rules), then:
+
+      1. Read the changed code (the coder result below names the files).
+         Look for edge cases the plan missed: empty inputs, error paths,
+         concurrency, partial failure, compat.
+      2. For each edge case belonging to a LATER step: check that step's
+         plan under {{plans_dir}}/steps/. If its Edge cases section already
+         covers it, done. If not, append an entry to that section via
+         fs_patch - touch NOTHING else in the file.
+      3. NEVER edit a later plan's Objective, Tasks, Acceptance criteria,
+         or Out of scope. Scope-affecting changes become proposed diffs in
+         `downstream_updates` instead.
+      4. Set needs_independent_review=true when the change touched 5+ files
+         or crosses architectural boundaries (auth, public APIs, schema,
+         security-sensitive paths).
+
+      Be terse. Findings, not prose.
+    prompt: |
+      ## Coder result
+      {{coder_result}}
+
+      ## Step plan
+      {{step_plan}}
+
+      ## Staleness report from orientation
+      {{staleness_report}}
+    tools:
+      - fs_cat
+      - fs_ls
+      - fs_patch
+      - execute_command
+    max_iterations: 20
+    output_schema:
+      type: object
+      properties:
+        edge_case_report:
+          type: string
+          description: Edge cases discovered - both handled and punted, one per line. "none" if empty
+        downstream_updates:
+          type: string
+          description: Annotations made (plan file + section) and proposed diffs for scope-affecting changes. "none" if empty
+        needs_independent_review:
+          type: boolean
+      required: [edge_case_report, downstream_updates, needs_independent_review]
+    fallback: write_handoff
+    next: route_sweep
+
+  route_sweep:
+    id: route_sweep
+    type: script
+    description: Broad or boundary-crossing changes get an independent reviewer.
+    script: scripts/route_sweep.sh
+    timeout: 5
+    fallback: write_handoff
+
+  independent_review:
+    id: independent_review
+    type: agent
+    description: Independent review pass - the author's self-review cannot catch its own rationalizations.
+    agent: code-reviewer
+    prompt: |
+      Review the changes produced for step {{step_number}} ({{step_title}})
+      of a phased implementation plan in {{project_dir}}.
+
+      What the step was supposed to do:
+      {{plan_summary}}
+
+      Coder summary (names the modified/created files):
+      {{coder_result}}
+
+      Review the changed files against the step plan's acceptance criteria.
+      Preserve severity tags in your findings.
+    timeout: 1200
+    state_updates:
+      review_report: '{{output}}'
+    next: route_review
+
+  route_review:
+    id: route_review
+    type: script
+    description: Critical findings loop back to implement (bounded); otherwise proceed to handoff.
+    script: scripts/route_review.sh
+    timeout: 5
+    fallback: write_handoff
+
+  write_handoff:
+    id: write_handoff
+    type: llm
+    description: |
+      Write the evidence-backed handoff per handoff-protocol and append
+      durable facts to NOTES.md. The completion gate (check_handoff)
+      verifies the document afterward.
+    skills_enabled: true
+    enabled_skills:
+      - handoff-protocol
+      - ai-slop-remover
+    instructions: |
+      Load `handoff-protocol` and follow its writer schema EXACTLY: the
+      frontmatter (step, title, result) and all eight sections, writing
+      "None" rather than omitting a section.
+
+      Write the handoff to {{handoff_path}} with fs_write. Paste the
+      verification evidence below verbatim into the Evidence section -
+      commands, exit codes, decisive output lines. Deviations come from the
+      staleness report, gate decisions, and fix loop history. Downstream
+      plan updates come from the sweep results.
+
+      VERIFICATION HONESTY: evidence marked "GATE NOT RUN" means that gate
+      is UNVERIFIED — record it as not run; never paraphrase a skipped gate
+      as covered, passing, or handled elsewhere. A handoff that admits an
+      unverified gate is correct; one that dresses it up as verified poisons
+      every downstream reader.
+
+      Then append durable, step-independent facts (if any) to {{notes_path}}
+      - create the file if missing, never rewrite existing entries.
+
+      If "Gate feedback" below is non-empty, a previous handoff attempt
+      failed validation - fix exactly what it lists.
+    prompt: |
+      ## Step
+      {{step_number}} ({{step_title}}) - plan at {{step_plan_path}}
+
+      ## Plan summary
+      {{plan_summary}}
+
+      ## Coder result
+      {{coder_result}}
+
+      ## Staleness report / deviations
+      {{staleness_report}}
+
+      Major deviation summary (if any): {{deviation_summary}}
+      User guidance given (if any): {{user_feedback}}
+      Fix loop attempts used: {{fix_attempts}} of {{max_fix_attempts}}
+
+      ## Edge cases discovered
+      {{edge_case_report}}
+
+      ## Downstream plan updates
+      {{downstream_updates}}
+
+      ## Independent review report (if any)
+      {{review_report}}
+
+      ## Verification evidence (paste verbatim)
+      ### Format
+      {{format_output}}
+      ### Lint
+      {{lint_output}}
+      ### Build
+      {{build_output}}
+      ### Tests
+      {{tests_output}}
+
+      ## Gate feedback
+      {{handoff_fix}}
+    tools:
+      - fs_cat
+      - fs_ls
+      - fs_write
+      - fs_patch
+    max_iterations: 15
+    output_schema:
+      type: object
+      properties:
+        step_summary:
+          type: string
+          description: 3-6 sentence summary of the step for the user's approval decision - what was done, deviations, anything needing their attention
+      required: [step_summary]
+    fallback: end_failure
+    next: check_handoff
+
+  check_handoff:
+    id: check_handoff
+    type: script
+    description: |
+      Deterministic completion gate - handoff exists with frontmatter and all
+      required sections. On success, marks the step plan status complete.
+      One retry back to write_handoff, then failure.
+    script: scripts/check_handoff.sh
+    timeout: 10
+    fallback: end_failure
+
+  gate_user_review:
+    id: gate_user_review
+    type: approval
+    description: The hard stop - the next step never starts without explicit approval.
+    question: |
+      ## Step {{step_number}} ({{step_title}}) - ready for review
+
+      {{step_summary}}
+
+      Handoff: {{handoff_path}}
+      Build: {{build_ok}} | Tests: {{tests_ok}} | Fix attempts: {{fix_attempts}}/{{max_fix_attempts}}
+
+      Approve this step? (Answer with anything else to send revision
+      instructions straight to the implementer.)
+    options:
+      - 'approve'
+      - 'revise'
+    routes:
+      'approve': end_success
+      'revise': get_revision
+    on_other: revise_from_choice
+    state_updates:
+      user_feedback: '{{choice}}'
+
+  get_revision:
+    id: get_revision
+    type: input
+    description: Collect revision instructions, then loop back through implement -> verify -> handoff.
+    question: 'What should change? Your comments go to the implementer verbatim.'
+    validation: 'len(input) > 0'
+    state_updates:
+      fix_instructions: '{{input}}'
+    next: implement
+
+  revise_from_choice:
+    id: revise_from_choice
+    type: script
+    description: Free-form approval answers are treated as revision instructions.
+    script: scripts/revise_from_choice.sh
+    timeout: 5
+    fallback: get_revision
+
+  end_success:
+    id: end_success
+    type: end
+    output: |
+      STEP_COMPLETE
+      Step: {{step_number}} ({{step_title}})
+      Plan: {{step_plan_path}}
+      Handoff: {{handoff_path}}
+      Build: passed | Tests: passed | Fix attempts: {{fix_attempts}}/{{max_fix_attempts}}
+
+      {{step_summary}}
+
+      Downstream plan updates:
+      {{downstream_updates}}
+
+  end_blocked:
+    id: end_blocked
+    type: end
+    output: |
+      STEP_BLOCKED
+      Step: {{step_number}} ({{step_title}})
+      Reason:
+      {{blocking_reason}}
+
+  end_rejected:
+    id: end_rejected
+    type: end
+    output: |
+      STEP_REJECTED
+      Step: {{step_number}} ({{step_title}})
+      Rejected at: deviation gate or coder approval gate.
+      Deviation summary:
+      {{deviation_summary}}
+      Coder result (if it ran):
+      {{coder_result}}
+
+  end_failure:
+    id: end_failure
+    type: end
+    output: |
+      STEP_FAILED
+      Step: {{step_number}} ({{step_title}})
+      Fix attempts: {{fix_attempts}}/{{max_fix_attempts}}
+      Blocking reason (if resolution failed): {{blocking_reason}}
+
+      Coder result:
+      {{coder_result}}
+
+      Last build output:
+      {{build_output}}
+
+      Last tests output:
+      {{tests_output}}

assets/agents/step-runner/scripts/check_handoff.sh

@@ -0,0 +1,54 @@
+#!/usr/bin/env bash
+set -uo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+handoff_path=$(echo "$state" | jq -r '.handoff_path // ""')
+step_plan_path=$(echo "$state" | jq -r '.step_plan_path // ""')
+handoff_attempts=$(echo "$state" | jq -r '.handoff_attempts // 0')
+
+problems=""
+
+if [[ ! -f "$handoff_path" ]]; then
+  problems="- handoff file does not exist at $handoff_path"$'\n'
+else
+  content=$(cat "$handoff_path")
+  grep -qE '^result:[[:space:]]*(complete|partial|blocked)' <<< "$content" \
+    || problems+="- frontmatter is missing 'result: complete|partial|blocked'"$'\n'
+  for section in "Summary" "Completed" "Not completed" "Deviations" "Downstream plan updates" "Edge cases discovered" "Evidence" "Notes for next step"; do
+    grep -qE "^##[[:space:]]+${section}" <<< "$content" \
+      || problems+="- missing required section: ## ${section}"$'\n'
+  done
+fi
+
+if [[ -z "$problems" ]]; then
+  if [[ -f "$step_plan_path" ]]; then
+    tmp=$(mktemp)
+    awk 'BEGIN{n=0} /^---[[:space:]]*$/{n++; print; next} n==1 && /^status:/{print "status: complete"; next} {print}' "$step_plan_path" > "$tmp" && mv "$tmp" "$step_plan_path"
+  fi
+  jq -nc '{"handoff_fix": "", "_next": "gate_user_review"}'
+  exit 0
+fi
+
+if (( handoff_attempts >= 1 )); then
+  jq -nc \
+    --arg br "Handoff failed validation twice. Problems:
+$problems" \
+    '{"blocking_reason": $br, "_next": "end_failure"}'
+  exit 0
+fi
+
+jq -nc \
+  --arg hf "The previous handoff attempt failed validation. Fix exactly these problems:
+$problems" \
+  '{
+    "handoff_attempts": 1,
+    "handoff_fix": $hf,
+    "_next": "write_handoff"
+  }'

assets/agents/step-runner/scripts/fix_loop_gate.sh

@@ -0,0 +1,60 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+fix_attempts=$(echo "$state" | jq -r '.fix_attempts // 0')
+max_fix_attempts=$(echo "$state" | jq -r '.max_fix_attempts // 2')
+lint_ok=$(echo "$state" | jq -r '.lint_ok | if . == null then "true" else (. | tostring) end')
+build_ok=$(echo "$state" | jq -r '.build_ok | if . == null then "true" else (. | tostring) end')
+tests_ok=$(echo "$state" | jq -r '.tests_ok | if . == null then "true" else (. | tostring) end')
+lint_output=$(echo "$state" | jq -r '.lint_output // ""')
+build_output=$(echo "$state" | jq -r '.build_output // ""')
+tests_output=$(echo "$state" | jq -r '.tests_output // ""')
+
+if (( fix_attempts >= max_fix_attempts )); then
+  jq -nc \
+    --argjson n "$fix_attempts" \
+    '{
+      "fix_attempts": $n,
+      "_next": "end_failure"
+    }'
+  exit 0
+fi
+
+next_attempts=$((fix_attempts + 1))
+
+if [[ "$lint_ok" != "true" ]]; then
+  stage="lint"
+  output="$lint_output"
+elif [[ "$build_ok" != "true" ]]; then
+  stage="build"
+  output="$build_output"
+elif [[ "$tests_ok" != "true" ]]; then
+  stage="full test suite"
+  output="$tests_output"
+else
+  stage="verification"
+  output="fix_loop_gate was reached but no failing stage was recorded. Re-run verification."
+fi
+
+fix_instructions=$(printf '## Fix loop status (step-level attempt %d of %d)\n\nThe implementation passed the coder'"'"'s internal checks but failed step-level verification at the %s stage.\n\nOutput:\n```\n%s\n```\n\nIdentify the minimal fix and apply it. Do not refactor. Regressions in untouched code caused by this change are in scope.' \
+  "$next_attempts" "$max_fix_attempts" "$stage" "$output")
+
+jq -nc \
+  --argjson n "$next_attempts" \
+  --arg 'fi' "$fix_instructions" \
+  '{
+    "fix_attempts": $n,
+    "fix_instructions": $fi,
+    "lint_ok": true,
+    "build_ok": true,
+    "tests_ok": true,
+    "_next": "implement"
+  }'

assets/agents/step-runner/scripts/resolve_step.sh

@@ -0,0 +1,152 @@
+#!/usr/bin/env bash
+set -uo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+fail() {
+  jq -nc --arg r "$1" '{"blocking_reason": $r, "_next": "end_failure"}'
+  exit 0
+}
+
+project_dir="${LLM_AGENT_VAR_PROJECT_DIR:-.}"
+project_dir=$(cd "$project_dir" 2>/dev/null && pwd) || fail "project_dir does not exist: $project_dir"
+
+plans_dir="${LLM_AGENT_VAR_PLANS_DIR:-plans}"
+[[ "$plans_dir" != /* ]] && plans_dir="$project_dir/$plans_dir"
+steps_dir="$plans_dir/steps"
+handoffs_dir="$plans_dir/handoffs"
+notes_path="$plans_dir/NOTES.md"
+
+[[ -d "$steps_dir" ]] || fail "No step plans directory at $steps_dir (expected <plans_dir>/steps/NN-<slug>.md)"
+
+frontmatter() {
+  awk '/^---[[:space:]]*$/{n++; next} n==1{print} n>=2{exit}' "$1"
+}
+
+fm_value() {
+  echo "$1" | grep -E "^$2:" | head -1 | sed -E "s/^$2:[[:space:]]*//" | sed -E 's/^["'"'"']|["'"'"']$//g'
+}
+
+step="${LLM_AGENT_VAR_STEP:-next}"
+if [[ "$step" == "next" ]]; then
+  prompt_step=$(echo "$state" | jq -r '.initial_prompt // ""' | grep -oiE 'step[[:space:]#:]*[0-9]+' | head -1 | grep -oE '[0-9]+' || true)
+  [[ -n "$prompt_step" ]] && step="$prompt_step"
+fi
+
+plan_file=""
+if [[ "$step" == "next" ]]; then
+  first_pending=""
+  while IFS= read -r f; do
+    st=$(fm_value "$(frontmatter "$f")" "status")
+    if [[ "$st" == "in-progress" ]]; then
+      plan_file="$f"
+      break
+    fi
+    [[ -z "$first_pending" && ( "$st" == "pending" || -z "$st" ) ]] && first_pending="$f"
+  done < <(find "$steps_dir" -maxdepth 1 -name '*.md' | sort)
+  [[ -z "$plan_file" ]] && plan_file="$first_pending"
+  [[ -z "$plan_file" ]] && fail "No in-progress or pending step plans in $steps_dir"
+else
+  [[ "$step" =~ ^[0-9]+$ ]] || fail "step must be a number or 'next'; got: $step"
+  padded=$(printf '%02d' "$((10#$step))")
+  plan_file=$(find "$steps_dir" -maxdepth 1 \( -name "${padded}-*.md" -o -name "${step}-*.md" \) | sort | head -1)
+  [[ -n "$plan_file" ]] || fail "No step plan matching step $step in $steps_dir"
+fi
+
+bn=$(basename "$plan_file" .md)
+num_part="${bn%%-*}"
+[[ "$num_part" =~ ^[0-9]+$ ]] || fail "Step plan filename must start with a number: $bn"
+step_number=$((10#$num_part))
+step_slug="${bn#*-}"
+
+fm=$(frontmatter "$plan_file")
+step_title=$(fm_value "$fm" "title")
+[[ -z "$step_title" ]] && step_title="$step_slug"
+
+deps=$(echo "$fm" | awk '/^depends_on:/{f=1; print; next} f && /^[[:space:]]*-/{print; next} f{exit}' | grep -oE '[0-9]+' || true)
+unsatisfied=""
+for dep in $deps; do
+  dep_padded=$(printf '%02d' "$((10#$dep))")
+  dep_handoff=$(find "$handoffs_dir" -maxdepth 1 \( -name "${dep_padded}-*.md" -o -name "${dep}-*.md" \) 2>/dev/null | sort | head -1)
+  if [[ -z "$dep_handoff" ]]; then
+    unsatisfied+="- step $dep: no handoff found (step not executed?)"$'\n'
+    continue
+  fi
+  dep_result=$(fm_value "$(frontmatter "$dep_handoff")" "result")
+  if [[ "$dep_result" != "complete" ]]; then
+    unsatisfied+="- step $dep: handoff result is '$dep_result' (not complete): $dep_handoff"$'\n'
+  fi
+done
+
+prev_handoff_path="(none)"
+prev_handoff="(none - this is the first step)"
+prev_file=""
+prev_num=0
+while IFS= read -r h; do
+  hn="${h##*/}"
+  hn="${hn%%-*}"
+  [[ "$hn" =~ ^[0-9]+$ ]] || continue
+  n=$((10#$hn))
+  if (( n < step_number && n >= prev_num )); then
+    prev_num=$n
+    prev_file="$h"
+  fi
+done < <(find "$handoffs_dir" -maxdepth 1 -name '*.md' 2>/dev/null | sort)
+if [[ -n "$prev_file" ]]; then
+  prev_handoff_path="$prev_file"
+  prev_handoff=$(head -c 16000 "$prev_file")
+fi
+
+notes="(none)"
+[[ -f "$notes_path" ]] && notes=$(head -c 8000 "$notes_path")
+
+step_plan=$(head -c 24000 "$plan_file")
+handoff_path="$handoffs_dir/$(basename "$plan_file")"
+
+tmp=$(mktemp)
+awk 'BEGIN{n=0} /^---[[:space:]]*$/{n++; print; next} n==1 && /^status:/{print "status: in-progress"; next} {print}' "$plan_file" > "$tmp" && mv "$tmp" "$plan_file"
+
+next_node="orient"
+blocking_reason=""
+if [[ -n "$unsatisfied" ]]; then
+  next_node="gate_blocked"
+  blocking_reason="Unsatisfied dependencies:"$'\n'"$unsatisfied"
+fi
+
+jq -nc \
+  --arg pd "$project_dir" \
+  --arg pl "$plans_dir" \
+  --argjson sn "$step_number" \
+  --arg ss "$step_slug" \
+  --arg st "$step_title" \
+  --arg spp "$plan_file" \
+  --arg sp "$step_plan" \
+  --arg php "$prev_handoff_path" \
+  --arg ph "$prev_handoff" \
+  --arg np "$notes_path" \
+  --arg no "$notes" \
+  --arg hp "$handoff_path" \
+  --arg br "$blocking_reason" \
+  --arg nx "$next_node" \
+  '{
+    "project_dir": $pd,
+    "plans_dir": $pl,
+    "step_number": $sn,
+    "step_slug": $ss,
+    "step_title": $st,
+    "step_plan_path": $spp,
+    "step_plan": $sp,
+    "prev_handoff_path": $php,
+    "prev_handoff": $ph,
+    "notes_path": $np,
+    "notes": $no,
+    "handoff_path": $hp,
+    "blocking_reason": $br,
+    "_next": $nx
+  }'

assets/agents/step-runner/scripts/revise_from_choice.sh

@@ -0,0 +1,27 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+feedback=$(echo "$state" | jq -r '.user_feedback // ""')
+
+if [[ -z "$feedback" ]]; then
+  jq -nc '{"_next": "get_revision"}'
+  exit 0
+fi
+
+fix_instructions=$(printf '## Revision requested by the user at the step approval gate\n\nAddress these comments with minimal edits, then the step re-verifies and the handoff is rewritten:\n\n%s' \
+  "$feedback")
+
+jq -nc \
+  --arg 'fi' "$fix_instructions" \
+  '{
+    "fix_instructions": $fi,
+    "_next": "implement"
+  }'

assets/agents/step-runner/scripts/route_coder_result.sh

@@ -0,0 +1,27 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+coder_result=$(echo "$state" | jq -r '.coder_result // ""')
+
+case "$coder_result" in
+  *CODER_COMPLETE*)
+    jq -nc '{"_next": "verify_format_lint"}'
+    ;;
+  *CODER_REJECTED*)
+    jq -nc '{"_next": "end_rejected"}'
+    ;;
+  *CODER_FAILED*)
+    jq -nc '{"blocking_reason": "coder fix-loop exhausted; see coder result", "_next": "end_failure"}'
+    ;;
+  *)
+    jq -nc '{"blocking_reason": "coder returned no recognizable sentinel (expected CODER_COMPLETE / CODER_REJECTED / CODER_FAILED)", "_next": "end_failure"}'
+    ;;
+esac

assets/agents/step-runner/scripts/route_review.sh

@@ -0,0 +1,38 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+review_report=$(echo "$state" | jq -r '.review_report // ""')
+review_attempts=$(echo "$state" | jq -r '.review_attempts // 0')
+max_review_attempts=$(echo "$state" | jq -r '.max_review_attempts // 1')
+
+if ! grep -qF "🔴" <<< "$review_report"; then
+  jq -nc '{"_next": "write_handoff"}'
+  exit 0
+fi
+
+if (( review_attempts >= max_review_attempts )); then
+  jq -nc '{"_next": "write_handoff"}'
+  exit 0
+fi
+
+next_review=$((review_attempts + 1))
+fix_instructions=$(printf '## Independent review findings (attempt %d of %d)\n\nAn independent reviewer flagged CRITICAL (🔴) findings. Address ONLY the 🔴 findings with minimal edits. Do not refactor unrelated code.\n\n%s' \
+  "$next_review" "$max_review_attempts" "$review_report")
+
+jq -nc \
+  --argjson n "$next_review" \
+  --arg 'fi' "$fix_instructions" \
+  '{
+    "review_attempts": $n,
+    "fix_instructions": $fi,
+    "needs_independent_review": false,
+    "_next": "implement"
+  }'

assets/agents/step-runner/scripts/route_staleness.sh

@@ -0,0 +1,23 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+has_major=$(echo "$state" | jq -r '.has_major_deviation // false')
+
+if [[ "${STEP_AUTOAPPROVE:-0}" == "1" ]]; then
+  jq -nc '{"_next": "implement"}'
+  exit 0
+fi
+
+if [[ "$has_major" == "true" ]]; then
+  jq -nc '{"_next": "gate_deviation"}'
+else
+  jq -nc '{"_next": "implement"}'
+fi

assets/agents/step-runner/scripts/route_sweep.sh

@@ -0,0 +1,23 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+needs_review=$(echo "$state" | jq -r '.needs_independent_review // false')
+
+if [[ "${STEP_SKIP_REVIEW:-0}" == "1" ]]; then
+  jq -nc '{"_next": "write_handoff"}'
+  exit 0
+fi
+
+if [[ "$needs_review" == "true" ]]; then
+  jq -nc '{"_next": "independent_review"}'
+else
+  jq -nc '{"_next": "write_handoff"}'
+fi

assets/agents/step-runner/scripts/verify_build.sh

@@ -0,0 +1,58 @@
+#!/usr/bin/env bash
+set -uo pipefail
+
+# shellcheck disable=SC1091
+source "$(dirname "$0")/../../.shared/utils.sh"
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+project_dir=$(echo "$state" | jq -r '.project_dir // "."')
+project_dir=$(resolve_gate_dir "$project_dir")
+
+if [[ -n "${BUILD_CMD:-}" ]]; then
+  cmd="$BUILD_CMD"
+else
+  project_info=$(detect_project "$project_dir")
+  cmd=$(echo "$project_info" | jq -r '.check // .build // ""')
+fi
+
+if [[ -z "$cmd" || "$cmd" == "null" ]]; then
+  jq -nc '{
+    "build_ok": true,
+    "build_output": "(GATE NOT RUN: no build/check command configured or detected. This is NOT evidence that the build passed — set BUILD_CMD, and never report the build as verified.)",
+    "_next": "verify_tests"
+  }'
+  exit 0
+fi
+
+exit_code=0
+output=$(cd "$project_dir" && eval "$cmd" 2>&1) || exit_code=$?
+
+if (( exit_code == 0 )); then
+  jq -nc \
+    --arg out "Ran: $cmd
+
+$output" \
+    '{
+      "build_ok": true,
+      "build_output": $out,
+      "_next": "verify_tests"
+    }'
+else
+  jq -nc \
+    --arg out "Ran: $cmd
+Exit code: $exit_code
+
+$output" \
+    '{
+      "build_ok": false,
+      "build_output": $out,
+      "_next": "fix_loop_gate"
+    }'
+fi

assets/agents/step-runner/scripts/verify_format_lint.sh

@@ -0,0 +1,84 @@
+#!/usr/bin/env bash
+set -uo pipefail
+
+# shellcheck disable=SC1091
+source "$(dirname "$0")/../../.shared/utils.sh"
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+project_dir=$(echo "$state" | jq -r '.project_dir // "."')
+project_dir=$(resolve_gate_dir "$project_dir")
+project_info=$(detect_project "$project_dir")
+project_type=$(echo "$project_info" | jq -r '.type // "unknown"')
+
+format_cmd="${FORMAT_CMD:-}"
+if [[ -z "$format_cmd" ]]; then
+  format_cmd=$(echo "$project_info" | jq -r '.fmt // ""')
+fi
+if [[ "$format_cmd" == "null" ]]; then format_cmd=""; fi
+
+if [[ -z "$format_cmd" ]]; then
+  format_output="(GATE NOT RUN: no format command configured or detected for project type '$project_type'. This is NOT evidence that formatting is clean. Set FORMAT_CMD to enable.)"
+else
+  fmt_rc=0
+  fmt_out=$(cd "$project_dir" && eval "$format_cmd" 2>&1) || fmt_rc=$?
+  format_output="Ran: $format_cmd
+Exit code: $fmt_rc
+
+$fmt_out"
+fi
+
+lint_cmd="${LINT_CMD:-}"
+if [[ -z "$lint_cmd" ]]; then
+  lint_cmd=$(echo "$project_info" | jq -r '.lint // ""')
+fi
+# The skip message must read as a WARNING, never a reassurance: the previous
+# wording ("linting is covered by the build/check command") was quoted
+# verbatim by workers as false evidence that linting passed
+if [[ -z "$lint_cmd" || "$lint_cmd" == "null" ]]; then
+  jq -nc \
+    --arg fo "$format_output" \
+    '{
+      "format_output": $fo,
+      "lint_ok": true,
+      "lint_output": "(GATE NOT RUN: no lint command configured or detected. This is NOT evidence that linting passed — set LINT_CMD or add a Taskfile lint target, and never report linting as covered.)",
+      "_next": "verify_build"
+    }'
+  exit 0
+fi
+
+lint_rc=0
+lint_out=$(cd "$project_dir" && eval "$lint_cmd" 2>&1) || lint_rc=$?
+
+if (( lint_rc == 0 )); then
+  jq -nc \
+    --arg fo "$format_output" \
+    --arg lo "Ran: $lint_cmd
+
+$lint_out" \
+    '{
+      "format_output": $fo,
+      "lint_ok": true,
+      "lint_output": $lo,
+      "_next": "verify_build"
+    }'
+else
+  jq -nc \
+    --arg fo "$format_output" \
+    --arg lo "Ran: $lint_cmd
+Exit code: $lint_rc
+
+$lint_out" \
+    '{
+      "format_output": $fo,
+      "lint_ok": false,
+      "lint_output": $lo,
+      "_next": "fix_loop_gate"
+    }'
+fi

assets/agents/step-runner/scripts/verify_tests.sh

@@ -0,0 +1,58 @@
+#!/usr/bin/env bash
+set -uo pipefail
+
+# shellcheck disable=SC1091
+source "$(dirname "$0")/../../.shared/utils.sh"
+
+if [[ -n "${GRAPH_STATE_FILE:-}" ]]; then
+  state=$(cat "$GRAPH_STATE_FILE")
+elif [[ -n "${GRAPH_STATE:-}" ]]; then
+  state="$GRAPH_STATE"
+else
+  state='{}'
+fi
+
+project_dir=$(echo "$state" | jq -r '.project_dir // "."')
+project_dir=$(resolve_gate_dir "$project_dir")
+
+if [[ -n "${TEST_CMD:-}" ]]; then
+  cmd="$TEST_CMD"
+else
+  project_info=$(detect_project "$project_dir")
+  cmd=$(echo "$project_info" | jq -r '.test // ""')
+fi
+
+if [[ -z "$cmd" || "$cmd" == "null" ]]; then
+  jq -nc '{
+    "tests_ok": true,
+    "tests_output": "(GATE NOT RUN: no test command configured or detected. This is NOT evidence that tests passed — set TEST_CMD, and never report the suite as green.)",
+    "_next": "edge_case_sweep"
+  }'
+  exit 0
+fi
+
+exit_code=0
+output=$(cd "$project_dir" && eval "$cmd" 2>&1) || exit_code=$?
+
+if (( exit_code == 0 )); then
+  jq -nc \
+    --arg out "Ran: $cmd
+
+$output" \
+    '{
+      "tests_ok": true,
+      "tests_output": $out,
+      "_next": "edge_case_sweep"
+    }'
+else
+  jq -nc \
+    --arg out "Ran: $cmd
+Exit code: $exit_code
+
+$output" \
+    '{
+      "tests_ok": false,
+      "tests_output": $out,
+      "_next": "fix_loop_gate"
+    }'
+fi

assets/functions/mcp.json

@@ -1,24 +1,13 @@
 {
   "mcpServers": {
     "github": {
-      "type": "stdio",
-      "command": "docker",
-      "args": [
-        "run",
-        "-i",
-        "--rm",
-        "-e",
-        "GITHUB_PERSONAL_ACCESS_TOKEN",
-        "ghcr.io/github/github-mcp-server"
-      ],
-      "env": {
-        "GITHUB_PERSONAL_ACCESS_TOKEN": "YOUR_GITHUB_TOKEN"
-      }
+      "type": "http",
+      "url": "https://api.githubcopilot.com/mcp"
     },
     "atlassian": {
       "type": "stdio",
       "command": "npx",
-      "args": ["-y", "mcp-remote@0.1.13", "https://mcp.atlassian.com/v1/mcp"]
+      "args": ["-y", "mcp-remote@latest", "https://mcp.atlassian.com/v1/mcp"]
     },
     "docker": {
       "type": "stdio",
@@ -29,6 +18,10 @@
       "type": "stdio",
       "command": "uvx",
       "args": ["duckduckgo-mcp-server"]
+    },
+    "iwe": {
+      "type": "stdio",
+      "command": "iwec"
     }
   }
 }

assets/functions/sbx-mixin.yaml

@@ -0,0 +1,44 @@
+schemaVersion: '2'
+kind: mixin
+name: built-in-tools
+description: >
+  Installs binaries and allows network domains required by Coyote's built-in
+  global tools and the default MCP server set. Auto-applied by Coyote's sbx
+  mixin discovery when running `coyote --sandbox`.
+
+permissions:
+  network:
+    allow:
+      # fetch_url_via_jina + jina reader fallback
+      - 'r.jina.ai'
+      # get_current_weather (.sh, .py, .ts)
+      - 'wttr.in'
+      # search_arxiv (the .sh tool still uses http://, so :80 is required until fixed)
+      - 'export.arxiv.org'
+      - 'export.arxiv.org:80'
+      # search_arxiv + search_wikipedia may follow DOI redirects
+      - 'doi.org'
+      # search_wikipedia
+      - 'en.wikipedia.org'
+      # search_wolframalpha
+      - 'api.wolframalpha.com'
+      # web_search_perplexity
+      - 'api.perplexity.ai'
+      # web_search_tavily
+      - 'api.tavily.com'
+      # send_twilio
+      - 'api.twilio.com'
+      # MCP: github (built-in mcp.json: api.githubcopilot.com)
+      - 'api.githubcopilot.com'
+      # MCP: atlassian (built-in mcp.json: mcp-remote -> mcp.atlassian.com)
+      - 'mcp.atlassian.com'
+      # MCP: ddg-search (built-in mcp.json: uvx duckduckgo-mcp-server)
+      - 'duckduckgo.com'
+      - 'html.duckduckgo.com'
+      - 'lite.duckduckgo.com'
+      # MCP: npx-based servers (mcp-remote) pull from npm
+      - 'registry.npmjs.org'
+      # MCP: docker server may pull images from common registries
+      - 'ghcr.io'
+      - 'registry-1.docker.io'
+      - 'auth.docker.io'

assets/functions/scripts/run-agent.py

@@ -32,7 +32,7 @@ def main():
     agent_data = parse_raw_data(raw_data)
 
     root_dir = "{config_dir}"
-    setup_env(root_dir, agent_func)
+    setup_env(root_dir, agent_func, raw_data)
 
     agent_tools_path = os.path.join(root_dir, "agents/{agent_name}/tools.py")
     run(agent_tools_path, agent_func, agent_data)
@@ -65,13 +65,14 @@ def parse_argv():
     return agent_func, agent_data
 
 
-def setup_env(root_dir, agent_func):
+def setup_env(root_dir, agent_func, raw_data):
     load_env(os.path.join(root_dir, ".env"))
     os.environ["LLM_ROOT_DIR"] = root_dir
     os.environ["LLM_AGENT_NAME"] = "{agent_name}"
     os.environ["LLM_AGENT_FUNC"] = agent_func
     os.environ["LLM_AGENT_ROOT_DIR"] = os.path.join(root_dir, "agents", "{agent_name}")
     os.environ["LLM_AGENT_CACHE_DIR"] = os.path.join(root_dir, "cache", "{agent_name}")
+    os.environ["LLM_AGENT_RAW_JSON"] = raw_data
 
 
 def load_env(file_path):

assets/functions/scripts/run-agent.sh

@@ -32,6 +32,7 @@ setup_env() {
     export LLM_AGENT_ROOT_DIR="$LLM_ROOT_DIR/agents/{agent_name}"
     export LLM_AGENT_CACHE_DIR="$LLM_ROOT_DIR/cache/{agent_name}"
     export LLM_PROMPT_UTILS_FILE="{prompt_utils_file}"
+    export LLM_AGENT_RAW_JSON="$agent_data"
 }
 
 load_env() {

assets/functions/scripts/run-agent.ts

@@ -11,7 +11,7 @@ async function main(): Promise<void> {
   const agentData = parseRawData(rawData);
 
   const configDir = "{config_dir}";
-  setupEnv(configDir, agentFunc);
+  setupEnv(configDir, agentFunc, rawData);
 
   const agentToolsPath = join(configDir, "agents", "{agent_name}", "tools.ts");
   await run(agentToolsPath, agentFunc, agentData);
@@ -48,13 +48,14 @@ function parseArgv(): { agentFunc: string; rawData: string } {
   return { agentFunc, rawData: agentData };
 }
 
-function setupEnv(configDir: string, agentFunc: string): void {
+function setupEnv(configDir: string, agentFunc: string, rawData: string): void {
   loadEnv(join(configDir, ".env"));
   process.env["LLM_ROOT_DIR"] = configDir;
   process.env["LLM_AGENT_NAME"] = "{agent_name}";
   process.env["LLM_AGENT_FUNC"] = agentFunc;
   process.env["LLM_AGENT_ROOT_DIR"] = join(configDir, "agents", "{agent_name}");
   process.env["LLM_AGENT_CACHE_DIR"] = join(configDir, "cache", "{agent_name}");
+  process.env["LLM_AGENT_RAW_JSON"] = rawData;
 }
 
 function loadEnv(filePath: string): void {

assets/functions/scripts/run-tool.py

@@ -32,7 +32,7 @@ def main():
     tool_data = parse_raw_data(raw_data)
 
     root_dir = "{root_dir}"
-    setup_env(root_dir)
+    setup_env(root_dir, raw_data)
 
     tool_path = "{tool_path}.py"
     run(tool_path, "run", tool_data)
@@ -65,11 +65,12 @@ def parse_argv():
     return tool_data
 
 
-def setup_env(root_dir):
+def setup_env(root_dir, raw_data):
     load_env(os.path.join(root_dir, ".env"))
     os.environ["LLM_ROOT_DIR"] = root_dir
     os.environ["LLM_TOOL_NAME"] = "{function_name}"
     os.environ["LLM_TOOL_CACHE_DIR"] = os.path.join(root_dir, "cache", "{function_name}")
+    os.environ["LLM_TOOL_RAW_JSON"] = raw_data
 
 
 def load_env(file_path):

assets/functions/scripts/run-tool.sh

@@ -29,6 +29,7 @@ setup_env() {
     export LLM_TOOL_NAME="{function_name}"
     export LLM_TOOL_CACHE_DIR="$LLM_ROOT_DIR/cache/{function_name}"
     export LLM_PROMPT_UTILS_FILE="{prompt_utils_file}"
+    export LLM_TOOL_RAW_JSON="$tool_data"
 }
 
 load_env() {

assets/functions/scripts/run-tool.ts

@@ -11,7 +11,7 @@ async function main(): Promise<void> {
   const toolData = parseRawData(rawData);
 
   const rootDir = "{root_dir}";
-  setupEnv(rootDir);
+  setupEnv(rootDir, rawData);
 
   const toolPath = "{tool_path}.ts";
   await run(toolPath, "run", toolData);
@@ -45,11 +45,12 @@ function parseArgv(): string {
   return toolData;
 }
 
-function setupEnv(rootDir: string): void {
+function setupEnv(rootDir: string, rawData: string): void {
   loadEnv(join(rootDir, ".env"));
   process.env["LLM_ROOT_DIR"] = rootDir;
   process.env["LLM_TOOL_NAME"] = "{function_name}";
   process.env["LLM_TOOL_CACHE_DIR"] = join(rootDir, "cache", "{function_name}");
+  process.env["LLM_TOOL_RAW_JSON"] = rawData;
 }
 
 function loadEnv(filePath: string): void {

assets/functions/tools/ast_grep.sh

@@ -0,0 +1,81 @@
+#!/usr/bin/env bash
+set -e
+
+# @describe Structural code search using AST patterns (ast-grep). Matches syntax trees, not text,
+# so it finds code regardless of formatting: function calls with any arguments, definitions, etc.
+# Use meta-variables in patterns: $NAME matches one AST node, $$$ matches zero or more nodes.
+# Patterns must be COMPLETE, valid AST nodes in the target language: 'fn $NAME($$$) { $$$ }'
+# matches Rust fn definitions (with body - 'fn $NAME($$$)' alone parses as nothing and matches
+# nothing), 'foo($$$)' matches all calls to foo, '$X.unwrap()' matches all unwrap calls.
+# Prefer this over fs_grep when searching for code STRUCTURE (calls, definitions, signatures);
+# use fs_grep for plain text, comments, or strings.
+
+# @option --pattern! The AST pattern to search for (must parse as valid code in the target language)
+# @option --lang The target language (e.g. rust, typescript, tsx, javascript, python, go, java, c, cpp, kotlin, swift, ruby, php, css, html, yaml, json). Strongly recommended; without it files of every supported language are scanned
+# @option --path The directory OR file to search in (defaults to current working directory)
+# @option --glob File glob to narrow the search (e.g. "src/**/*.rs", "!**/tests/**")
+
+# @env LLM_OUTPUT=/dev/stdout The output path
+
+MAX_RESULTS=100
+MAX_OUTPUT_BYTES=32768
+
+resolve_binary() {
+    if command -v ast-grep &>/dev/null; then
+        echo "ast-grep"
+        return 0
+    fi
+    if command -v sg &>/dev/null && sg --version 2>/dev/null | grep -qi 'ast-grep'; then
+        echo "sg"
+        return 0
+    fi
+    return 1
+}
+
+main() {
+    # shellcheck disable=SC2154
+    local pattern="$argc_pattern"
+    local lang="${argc_lang:-}"
+    local search_path="${argc_path:-.}"
+    local glob="${argc_glob:-}"
+
+    local bin
+    if ! bin=$(resolve_binary); then
+        printf 'ast-grep is not installed. Fall back to fs_grep for this search.\nTo enable structural search, install ast-grep:\n  cargo install ast-grep --locked\n  brew install ast-grep\n  npm i -g @ast-grep/cli\n' >> "$LLM_OUTPUT"
+        return 0
+    fi
+
+    if [[ ! -e "$search_path" ]]; then
+        echo "Error: path not found: $search_path" >> "$LLM_OUTPUT"
+        return 1
+    fi
+
+    local args=(run --pattern "$pattern" --color never --heading never)
+    [[ -n "$lang" ]] && args+=(--lang "$lang")
+    [[ -n "$glob" ]] && args+=(--globs "$glob")
+    args+=("$search_path")
+
+    local output exit_code=0
+    output=$("$bin" "${args[@]}" 2>&1) || exit_code=$?
+
+    if [[ -z "$output" ]]; then
+        echo "No structural matches found for: $pattern" >> "$LLM_OUTPUT"
+        return 0
+    fi
+
+    if (( exit_code > 1 )); then
+        printf 'ast-grep failed (exit %s):\n%s\n\nHint: the pattern must be valid %s syntax. Meta-variables: $NAME (one node), $$$ (zero or more).\n' \
+            "$exit_code" "$output" "${lang:-source}" >> "$LLM_OUTPUT"
+        return 0
+    fi
+
+    local total
+    total=$(wc -l <<< "$output")
+    output=$(head -n "$MAX_RESULTS" <<< "$output" | head -c "$MAX_OUTPUT_BYTES")
+
+    echo "$output" >> "$LLM_OUTPUT"
+    if (( total > MAX_RESULTS )); then
+        printf '\n(Showing %s of %s matching lines. Narrow with --glob, --lang, or a more specific pattern.)\n' \
+            "$MAX_RESULTS" "$total" >> "$LLM_OUTPUT"
+    fi
+}

assets/functions/tools/execute_command.sh

@@ -1,7 +1,7 @@
 #!/usr/bin/env bash
 set -e
 
-# @describe Execute the shell command.
+# @describe Execute the shell command. DO NOT use this to write files — use fs_write (new files) or fs_patch (edits) instead. Shell-based file writes (cat >, echo >, printf >, tee, heredocs, python -c "open(...)") break on multi-line content, special characters, quoted strings, and nested language blocks.
 # @option --command! The command to execute.
 
 # @env LLM_OUTPUT=/dev/stdout The output path
@@ -10,7 +10,15 @@ set -e
 source "$LLM_PROMPT_UTILS_FILE"
 
 main() {
-    guard_operation
     # shellcheck disable=SC2154
-    eval "$argc_command" >> "$LLM_OUTPUT"
+    argc_command="$(jq -r '.command' <<< "$LLM_TOOL_RAW_JSON")"
+
+    guard_operation
+    local script
+    script="$(mktemp)"
+    # shellcheck disable=SC2064
+    trap "rm -f '$script'" EXIT
+    # shellcheck disable=SC2154
+    printf '%s\n' "$argc_command" > "$script"
+    bash -e -o pipefail "$script" >> "$LLM_OUTPUT"
 }

assets/functions/tools/execute_sql_code.sh

@@ -14,6 +14,8 @@ source "$LLM_PROMPT_UTILS_FILE"
 
 # shellcheck disable=SC2154
 main() {
+    argc_code="$(jq -r '.code' <<< "$LLM_TOOL_RAW_JSON")"
+
     if ! grep -qi '^select' <<<"$argc_code"; then
         guard_operation ""
     fi

assets/functions/tools/fs_cat.sh

@@ -10,5 +10,13 @@ set -e
 
 main() {
     # shellcheck disable=SC2154
-		cat "$argc_path" >> "$LLM_OUTPUT" 2>&1 || echo "No such file or path: $argc_path" >> "$LLM_OUTPUT"
-}
+    local path="$argc_path"
+
+    # An empty result is shown to the model as the opaque literal "DONE"; emit a note instead.
+    if [[ -f "$path" && ! -s "$path" ]]; then
+        echo "(empty file: $path)" >> "$LLM_OUTPUT"
+        return 0
+    fi
+
+    cat "$path" >> "$LLM_OUTPUT" 2>&1 || echo "No such file or path: $path" >> "$LLM_OUTPUT"
+}

assets/functions/tools/fs_glob.sh

@@ -17,8 +17,8 @@ main() {
     local search_path="${argc_path:-.}"
 
     if [[ ! -d "$search_path" ]]; then
-        echo "Error: directory not found: $search_path" >> "$LLM_OUTPUT"
-        return 1
+        echo "Error: directory not found: $search_path" >&2
+        exit 1
     fi
 
     local results
@@ -33,7 +33,18 @@ main() {
             --exclude '.build' \
             2>/dev/null | head -n "$MAX_RESULTS") || true
     else
-        results=$(find "$search_path" -type f -name "$glob_pattern" \
+        local name_pattern dir_prefix effective_search
+        name_pattern="${glob_pattern##*/}"
+        [[ -z "$name_pattern" || "$name_pattern" == "**" ]] && name_pattern="*"
+        if [[ "$glob_pattern" == */* ]]; then
+            dir_prefix="${glob_pattern%%\**}"
+            dir_prefix="${dir_prefix%/}"
+            effective_search="${search_path}${dir_prefix:+/$dir_prefix}"
+        else
+            effective_search="$search_path"
+        fi
+        [[ -d "$effective_search" ]] || effective_search="$search_path"
+        results=$(find "$effective_search" -type f -name "$name_pattern" \
             -not -path '*/.git/*' \
             -not -path '*/node_modules/*' \
             -not -path '*/target/*' \

assets/functions/tools/fs_grep.sh

@@ -3,10 +3,11 @@ set -e
 
 # @describe Search file contents using regular expressions. Returns matching file paths and lines.
 # Use this to find relevant code before reading files. Much faster than reading files to search.
+# --path accepts either a directory (recursive search with exclude rules applied) or a single file.
 
 # @option --pattern! The regex pattern to search for in file contents
-# @option --path The directory to search in (defaults to current working directory)
-# @option --include File pattern to filter by (e.g. "*.rs", "*.{ts,tsx}", "*.py")
+# @option --path The directory OR file to search in (defaults to current working directory)
+# @option --include File pattern to filter by (e.g. "*.rs", "*.{ts,tsx}", "*.py"). Ignored when --path is a single file.
 
 # @env LLM_OUTPUT=/dev/stdout The output path
 
@@ -19,33 +20,39 @@ main() {
     local search_path="${argc_path:-.}"
     local include_filter="${argc_include:-}"
 
-    if [[ ! -d "$search_path" ]]; then
-        echo "Error: directory not found: $search_path" >> "$LLM_OUTPUT"
-        return 1
+    if [[ ! -e "$search_path" ]]; then
+        echo "Error: path not found: $search_path" >&2
+        exit 1
     fi
 
-    local grep_args=(-rn --color=never)
+    local grep_args=(-nH --color=never)
 
-    grep_args+=(
-        --exclude-dir='.git'
-        --exclude-dir='node_modules'
-        --exclude-dir='target'
-        --exclude-dir='dist'
-        --exclude-dir='build'
-        --exclude-dir='__pycache__'
-        --exclude-dir='vendor'
-        --exclude-dir='.build'
-        --exclude-dir='.next'
-        --exclude='*.min.js'
-        --exclude='*.min.css'
-        --exclude='*.map'
-        --exclude='*.lock'
-        --exclude='package-lock.json'
-    )
-
-    if [[ -n "$include_filter" ]]; then
-        grep_args+=("--include=$include_filter")
+    if [[ -d "$search_path" ]]; then
+        # Use -r (not -R) so symlinks to directories are NOT followed - this avoids
+        # infinite loops on pathological symlink cycles (e.g. `ln -s . loop`).
+        grep_args+=(-r)
+        grep_args+=(
+            --exclude-dir='.git'
+            --exclude-dir='node_modules'
+            --exclude-dir='target'
+            --exclude-dir='dist'
+            --exclude-dir='build'
+            --exclude-dir='__pycache__'
+            --exclude-dir='vendor'
+            --exclude-dir='.build'
+            --exclude-dir='.next'
+            --exclude='*.min.js'
+            --exclude='*.min.css'
+            --exclude='*.map'
+            --exclude='*.lock'
+            --exclude='package-lock.json'
+        )
+        if [[ -n "$include_filter" ]]; then
+            grep_args+=("--include=$include_filter")
+        fi
     fi
+    # If --path is a single file, --include and the exclude rules are ignored
+    # (they only matter when recursing into a directory tree).
 
     local results
     results=$(grep "${grep_args[@]}" -E "$search_pattern" "$search_path" 2>/dev/null | head -n "$MAX_RESULTS") || true

assets/functions/tools/fs_ls.sh

@@ -9,5 +9,18 @@ set -e
 
 main() {
     # shellcheck disable=SC2154
-    ls -1 "$argc_path" >> "$LLM_OUTPUT" 2>&1 || echo "No such path: $argc_path" >> "$LLM_OUTPUT"
-}
+    local path="$argc_path"
+    local output
+
+    if ! output=$(ls -1 "$path" 2>&1); then
+        echo "$output" >> "$LLM_OUTPUT"
+        return 0
+    fi
+
+    # An empty result is shown to the model as the opaque literal "DONE"; emit a note instead.
+    if [[ -z "$output" ]]; then
+        echo "(empty directory: $path)" >> "$LLM_OUTPUT"
+    else
+        echo "$output" >> "$LLM_OUTPUT"
+    fi
+}

assets/functions/tools/fs_patch.sh

@@ -1,11 +1,35 @@
 #!/usr/bin/env bash
 set -e
 
-# @describe Apply a patch to a file at the specified path.
-# This can be used to edit a file without having to rewrite the whole file.
+# @describe Apply a unified-diff patch to a file at the specified path. Use this for editing an existing file. It's the
+# PREFERRED way to modify a file. Prefer this over fs_write whenever the file already exists: it sends less data,
+# preserves unchanged content automatically, and is less prone to accidental data loss from full rewrites.
+# Use fs_write only when you are creating a new file or doing a complete rewrite where most of the content changes.
+#
+# CRITICAL — the patch is matched byte-for-byte. There is no fuzzy matching, no whitespace tolerance, and no context shift:
+# - Context lines (prefixed with a single space) and removed lines (prefixed with '-') must equal the file content exactly.
+#   If unsure, fs_cat the file first and copy the bytes verbatim into your patch.
+# - JSON-escape the contents string ONCE. Each literal backslash in the file becomes \\ in the JSON contents string. So a
+#   shell line containing s|\\"|"|g must appear in JSON as s|\\\\\"|\"|g — NOT s|\\\\\\\"|\\\"|g. Over-escaping backslashes
+#   is the most common cause of "unable to apply patch" failures, especially in files with sed/jq/regex pipelines or
+#   embedded Python with quoted strings.
+# - Hunks are applied in order; the first hunk that fails aborts the whole patch — later hunks are NOT attempted.
+# - Hunks anchor at the FIRST exact match of their context in the file, so include enough context to make each hunk
+#   unique. Hunks must appear in file order.
+# - Every hunk needs at least one context or removed line to anchor it; a hunk containing only additions is an error.
+# - Unrecognized lines inside a hunk are hard errors: every hunk line must start with ' ' (context), '-' (removal),
+#   or '+' (addition).
+# - If you've edited this file in earlier tool calls, fs_cat it again before composing the patch. A stale view of the file
+#   produces context lines that no longer match.
+# - On failure the error message names the failing hunk and shows the expected-vs-actual line. Fix that specific line and
+#   retry — do not blindly resend a near-identical patch.
+#
+# For files with heavy escaping (sed/jq/regex pipelines, shell with embedded heredocs, deeply quoted strings), prefer
+# fs_write over chained fs_patch hunks to replace the entire file with the full new contents (i.e. original content +
+# your changes).
 
 # @option --path! The path of the file to apply the patch to
-# @option --contents! The patch to apply to the file
+# @option --content! The patch to apply to the file
 
 # @env LLM_OUTPUT=/dev/stdout The output path
 
@@ -14,12 +38,31 @@ source "$LLM_PROMPT_UTILS_FILE"
 
 # shellcheck disable=SC2154
 main() {
+		# Command substitution strips *all* trailing newlines and `jq -r` appends one
+		# of its own, so read with `-j` and pin the real end of the content with a
+		# sentinel that is removed afterwards.
+		argc_contents="$(jq -j '.content' <<< "$LLM_TOOL_RAW_JSON"; printf x)"
+		argc_contents="${argc_contents%x}"
+    argc_path="$(jq -r '.path' <<< "$LLM_TOOL_RAW_JSON")"
+
     if [[ ! -f "$argc_path" ]]; then
         error "Unable to find the specified file: $argc_path"
         exit 1
     fi
 
-    new_contents="$(patch_file "$argc_path" <(printf "%s" "$argc_contents"))"
+		# Same sentinel guard on the patched result, otherwise the trailing newline
+		# is stripped again on the way back out. `rc` preserves patch_file's exit
+		# status so a failure still aborts under `set -e`.
+		new_contents="$(patch_file "$argc_path" <(printf "%s" "$argc_contents"); rc=$?; printf x; exit "$rc")"
+		new_contents="${new_contents%x}"
+
+		# awk newline-terminates every printed line, so patching a file that lacks
+		# a final newline would silently add one. Preserve the original file's
+		# final-newline state instead.
+		if [[ -n "$(tail -c 1 "$argc_path")" ]]; then
+			new_contents="${new_contents%$'\n'}"
+		fi
+
     printf "%s" "$new_contents" | git diff --no-index "$argc_path" - || true
 
     guard_operation "Apply changes?"

assets/functions/tools/fs_read.sh

@@ -1,13 +1,15 @@
 #!/usr/bin/env bash
 set -e
 
-# @describe Read a file with line numbers, offset, and limit. For directories, lists entries.
-# Prefer this over fs_cat for controlled reading. Use offset/limit to read specific sections.
+# @describe Read a TRUNCATED view of a file with line numbers, offset, and limit. For directories, lists entries.
+# IMPORTANT: This tool truncates output — lines over 2000 chars are cut off, and output is capped at 2000 lines by default.
+# If you need the FULL, untruncated contents of a file, use fs_cat instead.
+# Use this tool when you want line numbers, want to read a specific section via --offset/--limit, or are scanning a large file.
 # Use the grep tool to find specific content before reading, then read with offset to target the relevant section.
 
 # @option --path! The absolute path to the file or directory to read
-# @option --offset The line number to start reading from (1-indexed, default: 1)
-# @option --limit The maximum number of lines to read (default: 2000)
+# @option --offset <INT> The line number to start reading from (1-indexed, default: 1)
+# @option --limit <INT> The maximum number of lines to read (default: 2000)
 
 # @env LLM_OUTPUT=/dev/stdout The output path
 
@@ -21,8 +23,8 @@ main() {
     local limit="${argc_limit:-2000}"
 
     if [[ ! -e "$target" ]]; then
-        echo "Error: path not found: $target" >> "$LLM_OUTPUT"
-        return 1
+        echo "Error: path not found: $target" >&2
+        exit 1
     fi
 
     if [[ -d "$target" ]]; then
@@ -31,9 +33,20 @@ main() {
     fi
 
     local total_lines file_bytes
-    total_lines=$(wc -l < "$target" 2>/dev/null || echo 0)
+    # awk counts a final line that lacks a trailing newline; wc -l would undercount it by one.
+    total_lines=$(awk 'END { print NR }' "$target" 2>/dev/null || echo 0)
     file_bytes=$(wc -c < "$target" 2>/dev/null || echo 0)
 
+    if [[ "$total_lines" -eq 0 ]]; then
+        echo "(file is empty: $target)" >> "$LLM_OUTPUT"
+        return 0
+    fi
+
+    if [[ "$offset" -gt "$total_lines" ]]; then
+        echo "(offset $offset is past the end of the file, which has $total_lines lines)" >> "$LLM_OUTPUT"
+        return 0
+    fi
+
     if [[ "$file_bytes" -gt "$MAX_BYTES" ]] && [[ "$offset" -eq 1 ]] && [[ "$limit" -ge 2000 ]]; then
         {
             echo "Warning: Large file (${file_bytes} bytes, ${total_lines} lines). Showing first ${limit} lines."
@@ -46,7 +59,8 @@ main() {
 
     sed -n "${offset},${end_line}p" "$target" 2>/dev/null | {
         local line_num=$offset
-        while IFS= read -r line; do
+        # `|| [[ -n "$line" ]]` keeps the final line when the file has no trailing newline.
+        while IFS= read -r line || [[ -n "$line" ]]; do
             if [[ ${#line} -gt $MAX_LINE_LENGTH ]]; then
                 line="${line:0:$MAX_LINE_LENGTH}... (truncated)"
             fi

assets/functions/tools/fs_write.sh

@@ -1,10 +1,12 @@
 #!/usr/bin/env bash
 set -e
 
-# @describe Write the full file contents to a file at the specified path.
+# @describe Write the FULL file contents to a file at the specified path. Use this for NEW files or COMPLETE rewrites
+# only. For editing an existing file, prefer fs_patch. It's a surgical edit that preserves unchanged content, requires
+# sending less data, and is less prone to accidental data loss.
 
 # @option --path! The path of the file to write to
-# @option --contents! The full contents to write to the file
+# @option --content! The full contents to write to the file
 
 # @env LLM_OUTPUT=/dev/stdout The output path
 
@@ -13,6 +15,14 @@ source "$LLM_PROMPT_UTILS_FILE"
 
 # shellcheck disable=SC2154
 main() {
+		# Command substitution strips *all* trailing newlines and `jq -r` appends one
+		# of its own, so read with `-j` and pin the real end of the content with a
+		# sentinel that is removed afterwards. Without this every written file loses
+		# its final newline, which breaks formatters such as `cargo fmt --check`.
+		argc_contents="$(jq -j '.content' <<< "$LLM_TOOL_RAW_JSON"; printf x)"
+		argc_contents="${argc_contents%x}"
+    argc_path="$(jq -r '.path' <<< "$LLM_TOOL_RAW_JSON")"
+
     if [[ -f "$argc_path" ]]; then
         printf "%s" "$argc_contents" | git diff --no-index "$argc_path" - || true
         guard_operation "Apply changes?"

assets/functions/tools/send_mail.sh

@@ -14,6 +14,10 @@ set -e
 
 # shellcheck disable=SC2154
 main() {
+    argc_recipient="$(jq -r '.recipient' <<< "$LLM_TOOL_RAW_JSON")"
+    argc_subject="$(jq -r '.subject' <<< "$LLM_TOOL_RAW_JSON")"
+    argc_body="$(jq -r '.body' <<< "$LLM_TOOL_RAW_JSON")"
+
     sender_name="${EMAIL_SENDER_NAME:-$(echo "$EMAIL_SMTP_USER" | awk -F'@' '{print $1}')}"
     printf "%s\n" "From: $sender_name <$EMAIL_SMTP_USER>
 To: $argc_recipient 

assets/functions/tools/web_search_coyote.sh

@@ -6,11 +6,11 @@ set -e
 
 # @option --query! The search query.
 
-# @meta require-tools loki
+# @meta require-tools coyote
 
 # @env WEB_SEARCH_MODEL=gemini:gemini-2.5-flash The model for web-searching.
 #
-# supported loki models:
+# supported coyote models:
 #   - gemini:gemini-2.0-*
 #   - vertexai:gemini-*
 #   - perplexity:*
@@ -22,15 +22,15 @@ main() {
     client="${WEB_SEARCH_MODEL%%:*}"
 
     if [[ "$client" == "gemini" ]]; then
-        export LOKI_PATCH_GEMINI_CHAT_COMPLETIONS='{".*":{"body":{"tools":[{"google_search":{}}]}}}'
+        export COYOTE_PATCH_GEMINI_CHAT_COMPLETIONS='{".*":{"body":{"tools":[{"google_search":{}}]}}}'
     elif [[ "$client" == "vertexai" ]]; then
-        export LOKI_PATCH_VERTEXAI_CHAT_COMPLETIONS='{
+        export COYOTE_PATCH_VERTEXAI_CHAT_COMPLETIONS='{
     "gemini-1.5-.*":{"body":{"tools":[{"googleSearchRetrieval":{}}]}},
     "gemini-2.0-.*":{"body":{"tools":[{"google_search":{}}]}}
 }'
     elif [[ "$client" == "ernie" ]]; then
-        export LOKI_PATCH_ERNIE_CHAT_COMPLETIONS='{".*":{"body":{"web_search":{"enable":true}}}}'
+        export COYOTE_PATCH_ERNIE_CHAT_COMPLETIONS='{".*":{"body":{"web_search":{"enable":true}}}}'
     fi
 
-    loki -m "$WEB_SEARCH_MODEL" "$argc_query" >> "$LLM_OUTPUT"
+    coyote -m "$WEB_SEARCH_MODEL" "$argc_query" >> "$LLM_OUTPUT"
 }

assets/functions/utils/prompt-utils.sh

@@ -186,7 +186,9 @@ input() {
 }
 
 confirm() {
-  trap "stty echo; exit" EXIT
+  # stty targets stdin, which is /dev/null when the host spawns tool scripts;
+  # point it at the real terminal and stay quiet when there isn't one.
+  trap "stty echo </dev/tty 2>/dev/null; exit" EXIT
   _prompt_text "$1 (y/N)"
   echo -en "\033[36m\c " >&2
 
@@ -229,7 +231,7 @@ list() {
   declare first_row
   first_row=$((last_row - opts_count + 1))
 
-  trap "_cursor_blink_on; stty echo; exit" 2
+  trap "_cursor_blink_on; stty echo </dev/tty 2>/dev/null; exit" 2
 
   _cursor_blink_off
 
@@ -275,7 +277,7 @@ checkbox() {
   declare first_row
   first_row=$((last_row - opts_count + 1))
 
-  trap "_cursor_blink_on; stty echo; exit" 2
+  trap "_cursor_blink_on; stty echo </dev/tty 2>/dev/null; exit" 2
 
   _cursor_blink_off
 
@@ -403,7 +405,7 @@ range() {
   declare current_row
   current_row=$((first_row - 1))
 
-  trap "_cursor_blink_on; stty echo; exit" 2
+  trap "_cursor_blink_on; stty echo </dev/tty 2>/dev/null; exit" 2
 
   _cursor_blink_off
 
@@ -506,16 +508,16 @@ open_link() {
 }
 
 guard_operation() {
-  if [[ -t 1 ]]; then
-  	if [[ -z "$AUTO_CONFIRM" && -z "$LLM_AGENT_VAR_AUTO_CONFIRM" ]]; then
-			ans="$(confirm "${1:-Are you sure you want to continue?}")"
+	if [[ -z "$AUTO_CONFIRM" && -z "$LLM_AGENT_VAR_AUTO_CONFIRM" ]]; then
+		# 2>/dev/tty: keep the prompt off the host-captured stderr pipe so it
+		# can't leak into tool_call_error JSON when the wrapped command fails.
+		ans="$(confirm "${1:-Are you sure you want to continue?}" 2>/dev/tty)"
 
-			if [[ "$ans" == 0 ]]; then
-				error "Operation aborted!" 2>&1
-				exit 1
-			fi
+		if [[ "$ans" == 0 ]]; then
+			error "Operation aborted!" 2>&1
+			exit 1
 		fi
-  fi
+	fi
 }
 
 # Here is an example of a patch block that can be applied to modify the file to request the user's name:
@@ -528,7 +530,11 @@ guard_operation() {
 # +    print(f"Hello {name}")
 patch_file() {
   awk '
-    FNR == NR {
+    function isHeaderPair(i) {
+      return (patchLines[i] ~ /^--- / && patchLines[i+1] ~ /^\+\+\+ / && patchLines[i+2] ~ /^@@/)
+    }
+
+    FILENAME == ARGV[1] {
       lines[FNR] = $0
       next;
     }
@@ -547,12 +553,7 @@ patch_file() {
       while (patchLineIndex <= totalPatchLines) {
         line = patchLines[patchLineIndex]
 
-        if (line ~ /^--- / || line ~ /^\+\+\+ /) {
-          patchLineIndex++
-          continue
-        }
-
-        if (line ~ /^@@ /) {
+        if (line ~ /^@@/) {
           mode = "hunk"
           hunkIndex++
           patchLineIndex++
@@ -560,7 +561,22 @@ patch_file() {
         }
 
         if (mode == "hunk") {
-          while (patchLineIndex <= totalPatchLines && line ~ /^[-+ ]|^\s*$/ && line !~ /^--- /) {
+          while (patchLineIndex <= totalPatchLines) {
+            line = patchLines[patchLineIndex]
+
+            if (line ~ /^\\ No newline/) {
+              patchLineIndex++
+              continue
+            }
+
+            if (isHeaderPair(patchLineIndex)) {
+              break
+            }
+
+            if (line !~ /^[-+ ]/ && line !~ /^[ \t]*$/) {
+              break
+            }
+
             sanitizedLine = substr(line, 2)
 
             if (line !~ /^\+/) {
@@ -574,21 +590,79 @@ patch_file() {
             }
 
             patchLineIndex++
-            line = patchLines[patchLineIndex]
           }
 
           mode = "none"
-        } else {
-          patchLineIndex++
+          continue
         }
+
+        if (isHeaderPair(patchLineIndex)) {
+          patchLineIndex += 2
+          continue
+        }
+
+        if (line ~ /^\\ No newline/) {
+          patchLineIndex++
+          continue
+        }
+
+        if (hunkIndex == 0) {
+          # Preamble before the first hunk: tolerate prose, code fences, and lone headers.
+          patchLineIndex++
+          continue
+        }
+
+        if (line ~ /^[ \t]*$/ || line ~ /^```/) {
+          patchLineIndex++
+          continue
+        }
+
+        print "error: unrecognized line in patch (line " patchLineIndex "): " line > "/dev/stderr"
+        print "" > "/dev/stderr"
+        print "Every line inside a hunk must start with \" \" (context), \"-\" (removal), or \"+\" (addition)." > "/dev/stderr"
+        exit 1
       }
 
       if (hunkIndex == 0) {
         print "error: no patch" > "/dev/stderr"
+        print "" > "/dev/stderr"
+        print "No hunk header was found. Each hunk must start with a line beginning \"@@\"" > "/dev/stderr"
+        print "(for example \"@@ ... @@\" or \"@@ -1,4 +1,4 @@\"). Inside a hunk, context lines" > "/dev/stderr"
+        print "start with a single space, removed lines with \"-\", and added lines with \"+\"." > "/dev/stderr"
         exit 1
       }
 
       totalHunks = hunkIndex
+
+      if (totalLines == 0) {
+        for (h = 1; h <= totalHunks; h++) {
+          if (hunkTotalOriginalLines[h] > 0) {
+            print "error: unable to apply patch" > "/dev/stderr"
+            print "" > "/dev/stderr"
+            print "Hunk " h " expects existing content but the file is empty." > "/dev/stderr"
+            exit 1
+          }
+        }
+
+        for (h = 1; h <= totalHunks; h++) {
+          for (i = 1; i <= hunkTotalUpdatedLines[h]; i++) {
+            print hunkUpdatedLines[h,i]
+          }
+        }
+
+        exit 0
+      }
+
+      for (h = 1; h <= totalHunks; h++) {
+        if (hunkTotalOriginalLines[h] == 0) {
+          print "error: unable to apply patch" > "/dev/stderr"
+          print "" > "/dev/stderr"
+          print "Hunk " h " contains no context or removed lines; include at least one" > "/dev/stderr"
+          print "context line so the hunk can be anchored." > "/dev/stderr"
+          exit 1
+        }
+      }
+
       hunkIndex = 1
 
       for (lineIndex = 1; lineIndex <= totalLines; lineIndex++) {
@@ -599,7 +673,15 @@ patch_file() {
           nextLineIndex = lineIndex + 1
 
           for (i = 2; i <= hunkTotalOriginalLines[hunkIndex]; i++) {
-            if (lines[nextLineIndex] != hunkOriginalLines[hunkIndex,i]) {
+            if (nextLineIndex > totalLines || lines[nextLineIndex] != hunkOriginalLines[hunkIndex,i]) {
+              if (i - 1 > bestPartialLen[hunkIndex]) {
+                bestPartialLen[hunkIndex] = i - 1
+                bestPartialAnchorLine[hunkIndex] = lineIndex
+                bestPartialHunkPos[hunkIndex] = i
+                bestPartialDivergeLine[hunkIndex] = nextLineIndex
+                bestPartialExpected[hunkIndex] = hunkOriginalLines[hunkIndex,i]
+                bestPartialActual[hunkIndex] = lines[nextLineIndex]
+              }
               nextLineIndex = 0
               break
             }
@@ -621,7 +703,36 @@ patch_file() {
         }
 
         if (hunkIndex != totalHunks + 1) {
+          failingHunk = hunkIndex
           print "error: unable to apply patch" > "/dev/stderr"
+          print "" > "/dev/stderr"
+          print "Hunk " failingHunk " of " totalHunks " did not match the file." > "/dev/stderr"
+
+          if (bestPartialLen[failingHunk] == 0) {
+            print "" > "/dev/stderr"
+            print "The first context/removed line of hunk " failingHunk " was not found anywhere in the file:" > "/dev/stderr"
+            print "  expected: " hunkOriginalLines[failingHunk, 1] > "/dev/stderr"
+          } else {
+            print "" > "/dev/stderr"
+            print "Closest match: anchored at file line " bestPartialAnchorLine[failingHunk] ", matched " bestPartialLen[failingHunk] " of " hunkTotalOriginalLines[failingHunk] " original lines before diverging." > "/dev/stderr"
+            print "" > "/dev/stderr"
+            if (bestPartialDivergeLine[failingHunk] > totalLines) {
+              print "The hunk expects additional lines beyond the end of the file (file has " totalLines " lines)." > "/dev/stderr"
+            } else {
+              print "At file line " bestPartialDivergeLine[failingHunk] " (hunk original line " bestPartialHunkPos[failingHunk] "):" > "/dev/stderr"
+              print "  expected: " bestPartialExpected[failingHunk] > "/dev/stderr"
+              print "  actual:   " bestPartialActual[failingHunk] > "/dev/stderr"
+            }
+          }
+
+          print "" > "/dev/stderr"
+          print "Lines must match byte-for-byte (no fuzzy matching). Check escaping, whitespace, and quoting." > "/dev/stderr"
+
+          if (failingHunk < totalHunks) {
+            print "" > "/dev/stderr"
+            print (totalHunks - failingHunk) " subsequent hunk(s) were not attempted (patcher aborts on first failure)." > "/dev/stderr"
+          }
+
           exit 1
         }
     }
@@ -655,19 +766,18 @@ guard_path() {
     exit 1
   fi
 
-  if [[ -t 1 ]]; then
-    path="$(_to_real_path "$1")"
-    confirmation_prompt="$2"
+	path="$(_to_real_path "$1")"
+	confirmation_prompt="$2"
 
-    if [[ ! "$path" == "$(pwd)"* && -z "$AUTO_CONFIRM" && -z "$LLM_AGENT_VAR_AUTO_CONFIRM" ]]; then
-			ans="$(confirm "$confirmation_prompt")"
+	if [[ ! "$path" == "$(pwd)"* && -z "$AUTO_CONFIRM" && -z "$LLM_AGENT_VAR_AUTO_CONFIRM" ]]; then
+		# 2>/dev/tty: see guard_operation — prevents prompt text leaking via captured stderr.
+		ans="$(confirm "$confirmation_prompt" 2>/dev/tty)"
 
-			if [[ "$ans" == 0 ]]; then
-				error "Operation aborted!" >&2
-				exit 1
-			fi
-    fi
-  fi
+		if [[ "$ans" == 0 ]]; then
+			error "Operation aborted!" >&2
+			exit 1
+		fi
+	fi
 }
 
 _to_real_path() {

assets/roles/create-prompt.md

@@ -1,3 +1,6 @@
+---
+skills_enabled: false
+---
 As a professional Prompt Engineer, your role is to create effective and innovative prompts for interacting with AI models.
 
 Your core skills include:

assets/roles/create-title.md

@@ -1,3 +1,6 @@
+---
+skills_enabled: false
+---
 Create a concise, 3-6 word title.
 
 **Notes**:

assets/roles/diagnose.md

@@ -0,0 +1,96 @@
+---
+name: diagnose
+enabled_tools:
+  - execute_command
+  - fs_cat
+  - fs_ls
+  - web_search_coyote
+skills_enabled: false
+auto_continue: true
+max_auto_continues: 10
+---
+You are an expert systems troubleshooter: equal parts SRE, sysadmin, network engineer, and homelab tinkerer. Your job
+is to diagnose and fix technical problems of any kind: services that won't start, networking failures, container
+issues, driver problems, permission errors, misbehaving hardware, broken configs, or anything else. You are not limited
+to code.
+
+<system>
+os: {{__os__}}
+distro: {{__os_distro__}}
+arch: {{__arch__}}
+shell: {{__shell__}}
+cwd: {{__cwd__}}
+now: {{__now__}}
+</system>
+
+## Prime Directive
+
+**You run the diagnostics yourself.** Never tell the user to run a command and paste the output back. Use the
+`execute_command` tool to gather evidence directly, then interpret the results for them. The user should watch you
+work, not act as your terminal.
+
+## Diagnostic Loop
+
+Work the loop until the problem is solved or genuinely blocked:
+
+1. **Reproduce & observe.** Run the failing thing (or inspect its state) to see the actual error with your own eyes.
+   Never diagnose from the user's paraphrase alone.
+2. **Establish what changed.** Most breakage follows a change: updates, config edits, reboots, new hardware, expired
+   certs/leases. Check timestamps, package logs, and recent history early.
+3. **Check the dumb stuff first.** Is the service running? Is it enabled? Is the interface up? Is the disk full? Is
+   DNS resolving? Is the clock right? Cheap checks before deep theories.
+4. **Isolate by layer.** Split the problem space in half with each test:
+   - Networking: bottom-up — link → IP/DHCP → routing → DNS → transport → application.
+   - Software: process alive? → logs → config → dependencies/permissions → environment → binary itself.
+   - Containers: daemon → image → container state → logs → mounts/networks → host resources.
+5. **Hypothesize, then test.** State your current best hypothesis in one line before each test, and change ONE
+   variable at a time. If a test disproves the hypothesis, say so and pivot; don't quietly move on.
+6. **Fix the root cause, not the symptom.** A restart that "fixes" it without explanation is a data point, not a fix.
+7. **Verify.** After any fix, re-run the original failing operation and confirm it now works. No verification, no
+   victory declaration.
+
+## Evidence Gathering
+
+- Primary sources, in rough order of value: exit codes and stderr, service/app logs (`journalctl`, `docker logs`,
+  files under `/var/log`), kernel messages (`dmesg`), state inspection (`systemctl status`, `ip`, `ss`, `df`, `free`,
+  `lsblk`, `nmcli`, `docker ps/inspect`), then config files.
+- Make every command non-interactive and bounded: `--no-pager` for `journalctl`/`systemctl`, `-n`/`--since` to limit
+  log output, `timeout 10 ...` for anything that might hang, `-c` counts for `ping`. Never launch interactive TUIs
+  (top, htop, lazydocker itself) — use their batch/one-shot modes or underlying CLIs instead.
+- Prefer unprivileged commands. When root is genuinely required, say why and use `sudo` (the user may get a password
+  prompt in their terminal — that's expected).
+- Search the web for exact error strings (quoted, with software name and version) when an error is unfamiliar or
+  smells like a known bug or recent regression. Distro wikis, GitHub issues, and bug trackers beat guessing.
+
+## Safety Rules
+
+Commands fall into three tiers:
+
+1. **Read-only / inspection** (status, logs, listing, ping, dig, cat): run freely, no permission needed.
+2. **Reversible state changes** (restart a service, bounce an interface, recreate a container, edit a config after
+   backing it up): announce what you're about to do and why in one sentence, then do it. Back up any file before
+   modifying it (`cp file file.bak.$(date +%s)`).
+3. **Destructive or hard-to-reverse actions** (deleting data or volumes, formatting, `dd`, partitioning, package
+   removal, firewall flushes, forced resets): STOP and ask for explicit confirmation first, including the exact
+   command and a rollback plan. Never run these on your own judgment.
+
+Additional hard rules:
+
+- Never print or transmit secrets. If command output contains tokens, keys, or passwords, redact them in your response.
+- Never disable security controls (firewalls, SELinux/AppArmor, certificate validation) as a "fix" — at most as a
+  temporary, clearly-labeled isolation test, restored immediately after.
+- If the evidence points to failing hardware or risk of data loss, stop, say so plainly, and present options before
+  touching anything else.
+
+## When to Stop Gathering Evidence
+
+Once you have two or more independent pieces of evidence pointing to the same root cause, **stop gathering and deliver your diagnosis**. Do not add more verification steps to verify your verification. If you notice yourself thinking "let me just confirm one more thing" after you have already reached a conclusion, that is the signal to stop and explain the diagnosis instead. More data is not always better — a timely diagnosis with strong evidence beats an exhaustive audit.
+
+## Communication
+
+- Lead with what you found, not what you did. Then show the key evidence: the command and the relevant lines of its
+  output (trimmed — never dump walls of text).
+- When the problem is multi-step, keep a running todo list so the user can follow the investigation.
+- On resolution, close with a short summary: **root cause → fix applied → how it was verified → how to prevent it**.
+- If you're blocked (needs physical access, a password you don't have, a reboot decision), say exactly what you need
+  and what you'll do once you have it.

assets/roles/explain-shell.md

@@ -1,3 +1,6 @@
+---
+skills_enabled: false
+---
 Provide a terse, single sentence description of the given shell command.
 Describe each argument and option of the command.
 Provide short responses in about 80 words.

assets/roles/repo-analyzer.md

@@ -9,8 +9,8 @@ security/configuration settings. The analysis aims to ensure a thorough understa
 structured and operates, enabling the creation of new files, maintaining consistency with existing practices, and the
 potential implementation of best practices.
 
-Should the root directory contain a `LOKI.md` file, this was generated by Loki and should be used as a reference
-point for all analysis, style questions, etc.
+Should the root directory contain a `COYOTE.md` (or `AGENTS.md`/`CLAUDE.md`) file, this contains human-curated project
+instructions and should be used as a reference point for all analysis, style questions, etc.
 
 **Objective:** Enable the AI to thoroughly analyze a software repository, providing detailed insights and guidelines on
 all relevant aspects for understanding and potentially contributing to the project.

assets/roles/shell.md

@@ -1,3 +1,6 @@
+---
+skills_enabled: false
+---
 Provide only {{__shell__}} commands for {{__os_distro__}} without any description.
 Ensure the output is a valid {{__shell__}} command.
 If there is a lack of details, provide most logical solution.

assets/roles/slack.md

@@ -1,6 +1,5 @@
 ---
 enabled_mcp_servers: slack
-temperature: 0.2
 ---
 You are an expert Slack assistant designed to assist with Slack workspaces via the slack MCP server. 
 You can perform various tasks related to Slack, such as sending messages to channels, searching for messages, and 

assets/sbx-kit/files/home/.config/git/hooks/pre-commit

@@ -0,0 +1,21 @@
+#!/bin/sh
+if [ -z "$SSH_AUTH_SOCK" ]; then
+    echo "WARNING: [git-ssh-sign] no SSH agent — cannot sign commits"
+fi
+
+KEY=$(ssh-add -L 2>/dev/null | head -1)
+
+if [ -z "$KEY" ]; then
+    echo "WARNING: [git-ssh-sign] no keys in SSH agent — cannot sign commits"
+fi
+
+KEY_FILE="/home/agent/.config/git/signing_key.pub"
+mkdir -p "$(dirname "$KEY_FILE")"
+printf '%s\n' "$KEY" > "$KEY_FILE"
+EMAIL=$(git config user.email 2>/dev/null || echo "agent@sandbox.local")
+printf '%s %s\n' "$EMAIL" "$KEY" > "/home/agent/.config/git/allowed_signers"
+
+local_hook="$(git rev-parse --git-dir)/hooks/pre-commit"
+if [ -x "$local_hook" ]; then
+    exec "$local_hook" "$@"
+fi

assets/sbx-kit/spec.yaml

@@ -0,0 +1,373 @@
+# Docker sbx agent kit for Coyote
+#
+# Setup (paths use $HOME so commands work in bash/zsh/PowerShell/Git Bash):
+#   sbx create --kit ./sbx-kit/ coyote --name testing .
+#   sbx cp $HOME/.config/coyote/ testing:/home/agent/.config/
+#   sbx run testing --kit ./sbx-kit/
+schemaVersion: '2'
+kind: sandbox
+name: coyote
+displayName: Coyote
+description: >
+  An all-in-one, batteries-included LLM CLI tool featuring Shell Assistant,
+  CLI & REPL mode, RAG, AI tools & agents, MCP servers, skills, and macros.
+
+sandbox:
+  image: 'darkalex17/coyote:v0.8.3'
+  entrypoint: ['bash', '-lc', 'exec /home/agent/.cargo/bin/coyote']
+
+permissions:
+  network:
+    allow:
+      # Coyote release + self-update + model-registry sync
+      - 'github.com'
+      - 'api.github.com'
+      - 'raw.githubusercontent.com'
+      - 'objects.githubusercontent.com'
+      - '*.githubusercontent.com'
+      # Package managers and developer tools (cargo, uv, pip — useful at runtime for user installs)
+      - 'crates.io'
+      - 'static.crates.io'
+      - 'pypi.org'
+      - 'files.pythonhosted.org'
+      - 'astral.sh'
+      - 'sh.rustup.rs'
+      - 'static.rust-lang.org'
+
+      # LLM model OAuth + API endpoints
+      - 'claude.ai'
+      - 'console.anthropic.com'
+      - 'accounts.google.com'
+      # *.googleapis.com covers oauth2 + userinfo + VertexAI regional endpoints
+      # (*-aiplatform.googleapis.com). Do not narrow without re-checking VertexAI.
+      - '*.googleapis.com'
+
+      # Bedrock and GitHub Models use signed / GitHub-PAT auth that the proxy
+      # cannot rewrite; credentials must be injected separately (see README
+      # "Extending"). NOTE: '*.amazonaws.com' matches exactly ONE label, so
+      # two-label regional Bedrock hosts must be enumerated explicitly
+      # ('**.' is declared but not yet enforced by sbx). Add your region
+      # via a mixin if it's missing below.
+      - '*.amazonaws.com'
+      - 'bedrock-runtime.us-east-1.amazonaws.com'
+      - 'bedrock-runtime.us-east-2.amazonaws.com'
+      - 'bedrock-runtime.us-west-2.amazonaws.com'
+      - 'bedrock-runtime.eu-west-1.amazonaws.com'
+      - 'bedrock-runtime.eu-central-1.amazonaws.com'
+      - 'bedrock-runtime.ap-southeast-2.amazonaws.com'
+      - 'bedrock-runtime.ap-northeast-1.amazonaws.com'
+      - 'models.inference.ai.azure.com'
+
+      # Proxy-managed LLM provider APIs. Every credentials[].apiKey.inject
+      # domain below MUST also appear here. sbx does not derive allow entries
+      # from inject rules.
+      - 'api.openai.com'
+      - 'api.anthropic.com'
+      - 'generativelanguage.googleapis.com'
+      - 'api.cohere.ai'
+      - 'api.groq.com'
+      - 'openrouter.ai'
+      - 'api.ai21.com'
+      - 'api.cloudflare.com'
+      - 'api.deepinfra.com'
+      - 'api.deepseek.com'
+      - 'api.mistral.ai'
+      - 'api.perplexity.ai'
+      - 'api.voyageai.com'
+      - 'api.x.ai'
+      - 'api.jina.ai'
+      - 'r.jina.ai'
+      - 'qianfan.baidubce.com'
+      - 'api.hunyuan.cloud.tencent.com'
+      - 'api.minimax.chat'
+      - 'api.moonshot.cn'
+      - 'dashscope.aliyuncs.com'
+      - 'open.bigmodel.cn'
+
+# Proxy-managed LLM providers: inside the sandbox each apiKey env var holds
+# the `proxy-managed` sentinel; the proxy injects the real value into the
+# request header per the inject rules at request time. Values are bound by
+# the user via credential bindings (`sbx secret set <service>`); Coyote
+# pre-seeds them from its vault at launch. Multiple domains may map to one
+# service (e.g. jina) so they share a single credential.
+credentials:
+  - service: openai
+    description: OpenAI API key, injected on api.openai.com
+    apiKey:
+      name: OPENAI_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.openai.com
+          scheme: bearer
+  - service: anthropic
+    description: Anthropic API key, injected as x-api-key on api.anthropic.com
+    apiKey:
+      name: ANTHROPIC_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.anthropic.com
+          header: x-api-key
+          format: '%s'
+  - service: gemini
+    description: Google Gemini API key, injected as x-goog-api-key on generativelanguage.googleapis.com
+    apiKey:
+      name: GEMINI_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: generativelanguage.googleapis.com
+          header: x-goog-api-key
+          format: '%s'
+  - service: cohere
+    description: Cohere API key, injected on api.cohere.ai
+    apiKey:
+      name: COHERE_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.cohere.ai
+          scheme: bearer
+  - service: groq
+    description: Groq API key, injected on api.groq.com
+    apiKey:
+      name: GROQ_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.groq.com
+          scheme: bearer
+  - service: openrouter
+    description: OpenRouter API key, injected on openrouter.ai
+    apiKey:
+      name: OPENROUTER_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: openrouter.ai
+          scheme: bearer
+  - service: ai21
+    description: AI21 Labs API key, injected on api.ai21.com
+    apiKey:
+      name: AI21_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.ai21.com
+          scheme: bearer
+  - service: cloudflare
+    description: Cloudflare Workers AI API key, injected on api.cloudflare.com
+    apiKey:
+      name: CLOUDFLARE_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.cloudflare.com
+          scheme: bearer
+  - service: deepinfra
+    description: DeepInfra API key, injected on api.deepinfra.com
+    apiKey:
+      name: DEEPINFRA_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.deepinfra.com
+          scheme: bearer
+  - service: deepseek
+    description: DeepSeek API key, injected on api.deepseek.com
+    apiKey:
+      name: DEEPSEEK_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.deepseek.com
+          scheme: bearer
+  - service: mistral
+    description: Mistral API key, injected on api.mistral.ai
+    apiKey:
+      name: MISTRAL_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.mistral.ai
+          scheme: bearer
+  - service: perplexity
+    description: Perplexity API key, injected on api.perplexity.ai
+    apiKey:
+      name: PERPLEXITY_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.perplexity.ai
+          scheme: bearer
+  - service: voyageai
+    description: Voyage AI API key, injected on api.voyageai.com
+    apiKey:
+      name: VOYAGE_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.voyageai.com
+          scheme: bearer
+  - service: xai
+    description: xAI (Grok) API key, injected on api.x.ai
+    apiKey:
+      name: XAI_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.x.ai
+          scheme: bearer
+  - service: jina
+    description: Jina API key, injected on api.jina.ai and r.jina.ai
+    apiKey:
+      name: JINA_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.jina.ai
+          scheme: bearer
+        - domain: r.jina.ai
+          scheme: bearer
+  - service: ernie
+    description: Baidu ERNIE API key, injected on qianfan.baidubce.com
+    apiKey:
+      name: ERNIE_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: qianfan.baidubce.com
+          scheme: bearer
+  - service: hunyuan
+    description: Tencent Hunyuan API key, injected on api.hunyuan.cloud.tencent.com
+    apiKey:
+      name: HUNYUAN_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.hunyuan.cloud.tencent.com
+          scheme: bearer
+  - service: minimax
+    description: MiniMax API key, injected on api.minimax.chat
+    apiKey:
+      name: MINIMAX_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.minimax.chat
+          scheme: bearer
+  - service: moonshot
+    description: Moonshot AI API key, injected on api.moonshot.cn
+    apiKey:
+      name: MOONSHOT_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: api.moonshot.cn
+          scheme: bearer
+  - service: qianwen
+    description: Alibaba Qianwen (DashScope) API key, injected on dashscope.aliyuncs.com
+    apiKey:
+      name: DASHSCOPE_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: dashscope.aliyuncs.com
+          scheme: bearer
+  - service: zhipuai
+    description: Zhipu AI (GLM) API key, injected on open.bigmodel.cn
+    apiKey:
+      name: ZHIPUAI_API_KEY
+      proxyManaged: true
+      inject:
+        - domain: open.bigmodel.cn
+          scheme: bearer
+
+environment:
+  variables:
+    IS_SANDBOX: '1'
+    COYOTE_LOG_LEVEL: INFO
+    COYOTE_CONFIG_DIR: /home/agent/.config/coyote
+    EDITOR: nano
+    # Alias for the gemini credential: v2 apiKey supports a single env name
+    # (GEMINI_API_KEY above). Coyote also recognizes GOOGLE_API_KEY, so keep
+    # it set to the sentinel. Header injection happens per-domain regardless
+    # of which env var the app reads.
+    GOOGLE_API_KEY: proxy-managed
+
+setup:
+  files:
+    - path: /home/agent/.config/git/ssh-signing-key-command
+      mode: '0755'
+      description: Resolve the forwarded SSH agent key for Git SSH signing
+      content: |
+        #!/bin/sh
+        set -e
+
+        if [ -z "$SSH_AUTH_SOCK" ]; then
+            echo "WARNING: [git-ssh-sign] no SSH agent - cannot sign commits" >&2
+        fi
+
+        key=$(ssh-add -L 2>/dev/null | head -n 1)
+        if [ -z "$key" ]; then
+            echo "WARNING: [git-ssh-sign] no keys in SSH agent - cannot sign commits" >&2
+        fi
+
+        config_dir="$GIT_SSH_SIGN_CONFIG_DIR"
+        if [ -z "$config_dir" ]; then
+            config_dir="/home/agent/.config/git"
+        fi
+        mkdir -p "$config_dir"
+
+        email=$(git config user.email 2>/dev/null || printf '%s' "agent@sandbox.local")
+        printf '%s %s\n' "$email" "$key" > "$config_dir/allowed_signers"
+        printf 'key::%s\n' "$key"
+
+  install:
+    - command: |
+        git config --system gpg.format ssh
+        git config --system --unset-all user.signingKey || true
+        git config --system commit.gpgSign true
+        git config --system tag.gpgSign true
+        git config --system gpg.ssh.defaultKeyCommand /home/agent/.config/git/ssh-signing-key-command
+        git config --system gpg.ssh.allowedSignersFile /home/agent/.config/git/allowed_signers
+        if [ "$(git config --system --get core.hooksPath || true)" = "/home/agent/.config/git/hooks" ]; then
+            git config --system --unset-all core.hooksPath
+        fi
+      user: '0'
+      description: Configure SSH commit signing with a dynamic key command
+
+  startup:
+    - command:
+        [
+          'sh',
+          '-c',
+          'test -f "$HOME/.config/coyote/config.yaml" || coyote --info >/dev/null 2>&1 || true',
+        ]
+      user: '1000'
+      background: false
+      description: Bootstrap Coyote config directory on first sandbox start
+
+agentInstructions:
+  filename: COYOTE.md
+  content: |
+    ## Sandbox environment
+
+    You are running inside a Docker sandbox launched via `sbx run coyote`. The
+    user's project workspace is mounted at its absolute host path and is the
+    current working directory. `sudo` is passwordless; use it for system
+    package installs.
+
+    Coyote's configuration lives at `~/.config/coyote/` and logs at
+    `~/.cache/coyote/coyote.log`. Persistence is enabled, so config, sessions,
+    vault state, OAuth tokens, and installed tools survive sandbox restarts.
+
+    LLM provider credentials are forwarded by the sandbox HTTP proxy via
+    credential bindings. Coyote pre-seeds them from its vault at launch
+    (`sbx secret set <service>`); users can also bind values manually on the
+    host with `sbx secret set <service>` or `sbx secret import`. Recognized
+    services:
+
+      openai, anthropic, gemini, cohere, groq, openrouter, ai21,
+      cloudflare, deepinfra, deepseek, mistral, perplexity, voyageai,
+      xai, jina, ernie, hunyuan, minimax, moonshot, qianwen, zhipuai
+
+    Inside the sandbox the corresponding env vars (OPENAI_API_KEY, etc.)
+    hold the placeholder string `proxy-managed`; the proxy substitutes the
+    real value at request time. OAuth flows for Claude Pro/Max and Gemini
+    are also allow-listed.
+
+    Bedrock (AWS) and VertexAI (Google Cloud) use signed/OAuth-token requests
+    that the proxy cannot rewrite, so you must inject credentials yourself via
+    `sbx run --env AWS_ACCESS_KEY_ID=...` or a mixin kit that mounts a
+    service-account JSON. VertexAI regional endpoints are allow-listed via
+    `*.googleapis.com`. Bedrock runtime endpoints are allow-listed for
+    us-east-1/2, us-west-2, eu-west-1, eu-central-1, ap-southeast-2, and
+    ap-northeast-1 only; other regions need a mixin allow entry
+    (`bedrock-runtime.<region>.amazonaws.com`).
+
+    Useful first-run commands:
+    - `coyote --info`          # show config paths and resolved settings
+    - `coyote --list-secrets`  # initialise the local vault
+    - `coyote --authenticate <client>`  # OAuth flow (Claude Pro/Max, Gemini)

assets/skills/adversarial-review/SKILL.md

@@ -0,0 +1,79 @@
+---
+description: Adversarial plan-conformance review of an implementation against the task/plan it was supposed to satisfy. Verdict is CONFORMS or DIVERGES with acceptance-criterion-referenced complaints. Grants read-only filesystem access for ground-truth checks. Complements code-review (which judges code quality); this judges whether the code is the RIGHT code per the plan.
+enabled_tools: fs_read, fs_grep, fs_glob, fs_cat, fs_ls
+---
+You are an adversarial plan-conformance reviewer. A code-quality reviewer already asks "is this code good?" — you ask a different, harder question: **"is this the code the plan asked for, and ONLY that?"** You are hunting for the gap between what was specified and what was built. Assume the implementer drifted, cut a corner, or misread the plan until the diff proves otherwise. Your independence is the value: you have no stake in the implementation decisions and no reason to rationalize them.
+
+You review THE CHANGE against THE PLAN. You are given (a) the diff, (b) the task/plan it implements — its Objective, Tasks, and above all its **Acceptance criteria**. If the plan is missing, say so and stop: you cannot judge conformance without a spec.
+
+## The core discipline: map every acceptance criterion to evidence
+
+For EACH acceptance criterion in the plan, find the specific evidence in the diff that satisfies it, and classify:
+
+| Verdict per criterion | Meaning |
+|---|---|
+| ✅ **Met** | The diff contains code that observably satisfies this criterion, AND a test that will fail if it regresses. Cite the file:line. |
+| ⚠️ **Partial** | Some of the criterion is implemented but a case, path, or sub-requirement is missing. Name what's missing. |
+| ❌ **Unmet** | No code in the diff satisfies this criterion. The "dog that didn't bark." |
+| 🔀 **Diverged** | The diff implements something ADJACENT to the criterion but not it — different interface, different behavior, different data shape than specified. |
+
+A criterion with no corresponding test is at best ⚠️ Partial — "implemented but unverifiable" is not "met." An acceptance criterion is a promise of observable behavior; if nothing proves the behavior, the promise is unkept.
+
+## What to hunt for (adversarial checklist)
+
+### 1. Silently skipped criteria (the dog that didn't bark)
+Read the acceptance criteria list, then the diff. Every criterion with no matching change is a finding. Implementers under-deliver far more often by *omission* than by writing wrong code. The absent migration, the un-added error path, the criterion #4 that quietly became "out of scope" without anyone deciding that — these are your highest-value catches.
+
+### 2. Silent scope drift
+- **Scope creep:** code in the diff that no criterion or task asked for. New abstractions, refactors of untouched code, "while I was in here" changes. Flag it — the plan defined the scope, and the implementer doesn't get to redefine it unilaterally.
+- **Interface drift:** the plan named a symbol/signature/endpoint/column exactly (`RecordPurchase` using `ExternalTierID`, a `tier_id` column, a specific RPC). The diff uses a different name or shape. Even if the code works, it diverged from the contract other steps depend on.
+- **Approach substitution:** the plan (or a recorded decision) said "do X, not Y, because Z." The diff does Y. The implementer re-litigated a settled decision. Flag it with the plan's stated reason.
+
+### 3. Ground-truth verification (verify, don't trust the diff's self-description)
+The diff shows what changed, not whether it's correct against the codebase:
+- `fs_grep` every symbol the plan requires — confirm the diff actually introduced/changed it, spelled as specified.
+- `fs_read` around each hunk to confirm the change lands in the right place and the enclosing scope makes the criterion true (not just that a line matching the keyword appears).
+- `fs_grep` the callers of anything changed — a criterion is not met if the new behavior isn't actually reached.
+- Confirm tests exist AND target the criterion's behavior, not the implementation. A tautological test (`assert x.is_empty() || !x.is_empty()`) counts as no test.
+
+### 4. Out-of-scope violations
+If the plan has an "Out of scope" section, check the diff didn't touch those things. Touching explicitly-excluded surface is a divergence even if the code is fine.
+
+### 5. Downstream contract breakage
+If this change creates a surface a LATER step depends on (per the plan's dependency graph), verify the surface matches what those downstream steps will expect. A rename here that breaks step N+2's stated assumption is a divergence you catch now or pay for later.
+
+## Verdict format
+
+End with EXACTLY one of:
+
+```
+ADVERSARIAL_REVIEW: CONFORMS
+Criteria: N/N met (all with tests).
+<optional: 1-3 non-blocking observations>
+```
+
+```
+ADVERSARIAL_REVIEW: DIVERGES
+Criteria: X/N met, Y partial, Z unmet/diverged.
+Complaints:
+1. Acceptance criterion "<quote the criterion>" — <Unmet|Partial|Diverged> — <what the diff does or fails to do, with file:line> — <what would make it conform>
+2. Scope drift — <file:line> — <what was added that no criterion asked for> — remove or get it into scope
+3. ...
+```
+
+Every complaint MUST tie to a specific acceptance criterion (quoted) or a specific scope/interface/out-of-scope violation, and MUST cite file:line. "The implementation seems incomplete" is noise; `criterion "returns 429 after 3 failed attempts" — Unmet — retry.go has no attempt counter; the loop retries forever (retry.go:41) — add a bounded counter and a test asserting the 4th call returns 429` is signal.
+
+## Scope discipline (what you are NOT)
+
+- You are NOT the code-quality reviewer. Do not flag style, naming aesthetics, micro-optimizations, or "I'd have written it differently" unless it causes a criterion to be unmet. The `code-review` skill owns quality; you own conformance. If a quality issue is severe enough to break a criterion (a race that violates a correctness criterion), flag it as a conformance failure and note it's also a quality issue.
+- You do NOT rewrite the code or the plan. You produce a verdict and complaints; the implementer owns the fix.
+- If the plan itself is wrong (asks for something impossible or self-contradictory), that is a DIVERGES with a complaint that the plan is the root cause — do not paper over it by judging against a plan you silently corrected.
+- Three decisive divergences beat fifteen weak ones. If every criterion is a nitpick, the change probably CONFORMS — say so.
+
+## Anti-patterns
+
+- Rubber-stamping CONFORMS because the code "looks done" without mapping each criterion to evidence.
+- Judging code quality instead of plan conformance (that's the other reviewer's job).
+- Accepting a criterion as met with no test proving it.
+- Missing a silently-skipped criterion because you only reviewed what's IN the diff, never what's ABSENT.
+- Complaints with no criterion reference and no file:line.

assets/skills/ai-slop-remover/SKILL.md

@@ -0,0 +1,39 @@
+---
+description: Detect and remove AI slop from code and prose; produce output indistinguishable from a senior engineer's.
+---
+You are reviewing or generating content. Apply these standards strictly. The goal is output that reads like it was written by a competent human professional, not an AI.
+
+## Code
+
+**No useless comments.** A comment is useless if it restates the code:
+- BAD: `// Increment counter` above `counter += 1`
+- BAD: `/// Returns the user's name.` on `fn user_name() -> &str`
+- GOOD: Comments that explain a non-obvious WHY: a constraint, an invariant, a workaround for a specific bug, behavior that would surprise a reader.
+
+If removing a comment wouldn't confuse a future reader, the comment shouldn't exist.
+
+**No emojis** unless the user explicitly asked for them.
+
+**No defensive handling for impossible cases.** If a function only receives valid input from internal callers, don't pretend otherwise. Validate at system boundaries (user input, external APIs, file I/O); trust internal code.
+
+**No over-engineering for hypothetical futures.** Three similar lines of code is fine. Premature abstractions are worse than duplication.
+
+**No backwards-compatibility cruft for unreleased code.** If a function isn't called yet, just change it. Don't add `_unused` prefixes, "// removed" comments, or wrapper layers "for migration."
+
+**Names should be honest.** A function called `get_user` should not mutate state. A field called `count` should not be a function. A method that can fail should return `Result`, not panic.
+
+## Prose
+
+**No flattery.** Don't start with "Great question!" or "That's a really good idea!" Just respond.
+
+**No filler.** "It's important to note that" — delete. "Let me explain" — just explain. "I'll go ahead and" — just do it.
+
+**No status updates.** "I'm going to help you with that" — just help.
+
+**Match the user's terseness.** Brief user, brief reply. Detailed user, detailed reply.
+
+**No multi-paragraph docstrings.** One short line max. If the function needs paragraphs to explain, the function is doing too much.
+
+## When in doubt
+
+Ask: "Would a senior engineer write this in a code review or a Slack message?" If not, cut it.

assets/skills/code-review/SKILL.md

@@ -0,0 +1,124 @@
+---
+description: Conduct a thorough code review focused on correctness, clarity, tests, and footguns. Grants read-only filesystem access for inspecting code.
+enabled_tools: fs_read, fs_grep, fs_glob, fs_cat, fs_ls
+---
+You are reviewing code. Use the filesystem tools (`fs_read`, `fs_grep`, `fs_glob`, `fs_cat`, `fs_ls`) to inspect files. Apply this checklist in order; stop at the first category where you find substantial issues, since fixing those usually shifts the rest of the review.
+
+## Investigation workflow
+
+Before reviewing, build a mental model of the surrounding code:
+
+- `fs_ls` the directories that contain the changed files.
+- `fs_grep` for the symbols being added/modified to see existing callers and tests.
+- `fs_read` neighboring files in the same module to understand local conventions.
+- `fs_glob` for test files that might cover this area.
+
+A review without context is just a syntax check.
+
+## Reviewing a diff
+
+When you only see a hunk (not the whole file), the default context is sparse — usually 3 lines on either side. You see what changed but rarely the function signature, the caller, or the test. Read deliberately to recover what the diff omits.
+
+### Read around the hunk
+
+The `@@ -120,8 +120,12 @@` header gives you the line numbers in the old (`-`) and new (`+`) file. Read 20–40 lines around the hunk to see the enclosing function:
+
+```
+fs_read --path "src/auth.rs" --offset 110 --limit 40
+```
+
+You're recovering: the function signature, the return type, what unchanged portions do, and whether the hunk's logic fits its enclosing scope.
+
+### Read the callers of anything changed
+
+If a hunk changes a function's body or its signature, grep for the name to find callers and check whether the change ripples:
+
+```
+fs_grep --pattern "changed_function" --include "*.rs"
+```
+
+Skip the test files in this search; do the test sweep next.
+
+### Read the tests for the change
+
+Even if the diff doesn't touch test files, check whether tests exist for what's changing:
+
+```
+fs_grep --pattern "changed_function" --include "*_test.rs"
+fs_grep --pattern "changed_function" --include "tests/*"
+```
+
+Absence of tests for a changed function is itself a finding ("changes function X but no test references it; regressions won't be caught").
+
+### Diff-shaped issues to watch for
+
+These are review findings that only surface in a diff context, not in a whole-file read:
+
+- **Renames** (`diff --git a/old.rs b/new.rs`) — `fs_grep` for the old path to find imports that need updating but weren't.
+- **Signature changes** — verify all callers compile against the new signature. Compiler-checked languages catch some of this; dynamic languages don't.
+- **New code path without new tests** — usually a missing test. Flag it.
+- **Removed code with tests still present** — the tests probably need updating too.
+- **The "dog that didn't bark"** — what's obvious by its ABSENCE? A new field with no migration, a new error path with no test, a public API change with no changelog, a new config option with no documentation. Flag these as missing pieces, not as things to add later.
+
+### Scope discipline
+
+A diff review is a review of THE CHANGE, not the whole file:
+
+- Don't moralize about pre-existing code unless the diff makes it worse.
+- Don't suggest refactors outside the scope of the change. ("This whole module could be cleaner" is not actionable feedback on a 5-line patch.)
+- If you spot unrelated bugs while reading context, mention them briefly but separately: prefix with `Pre-existing, out of scope:` so the author knows which findings block their merge and which are FYI.
+- The author's job is to ship THIS change. Your job is to catch what's wrong with THIS change.
+
+## 1. Correctness
+
+- Does the change actually do what it claims? Does it solve the stated problem?
+- Edge cases: empty inputs, max sizes, concurrent access, error paths, partial failures.
+- Off-by-one errors, type confusion, null/None handling, integer overflow.
+- Race conditions and ordering assumptions across threads, async tasks, or distributed components.
+- Resource cleanup: file handles, locks, network connections, transactions.
+
+## 2. Tests
+
+- Do the tests test BEHAVIOR, not implementation? (Tests of `private_helper()` are usually a smell.)
+- Will they fail when the code regresses? Or are they tautological (e.g., `assert!(x.is_empty() || !x.is_empty())`)?
+- Do they cover the unhappy paths, not just the happy ones?
+- Is there a missing test for the specific bug or feature being added? `fs_grep` for the function name in test files to check.
+
+## 3. Clarity
+
+- Are names accurate? `get_user` that mutates is a lie; rename or split.
+- Could a competent reader understand this without comments?
+- Is there a simpler way to express the same logic?
+- Is the function doing one thing, or several things glued together?
+
+## 4. Coupling
+
+- Does this change increase coupling between modules unnecessarily?
+- Is the new code reaching into internals it shouldn't (private fields exposed, deep import paths)?
+- Could the change be expressed as a smaller diff that doesn't ripple through unrelated files?
+
+## 5. Footguns
+
+- Could a future maintainer easily misuse this API?
+- Are invariants enforced by types, or just by convention?
+- Are error types specific enough to be actionable?
+- Is there a documented or implicit ordering requirement that's easy to break?
+
+## What to flag
+
+- Correctness bugs.
+- Missing error handling at trust boundaries.
+- Race conditions.
+- Tests that won't catch regressions.
+- Security issues (injection, auth, exposed secrets).
+
+## What to let go
+
+- Style differences that aren't in the codebase's existing conventions.
+- "I would have done it differently" preferences.
+- Comments and naming choices that match existing patterns in the same file.
+- Micro-optimizations in code that isn't on a hot path.
+
+## Tone
+
+Direct, specific, focused on the code. No flattery, no padding. If something is wrong, say so plainly with the file path and line reference and the reason. If something is good and non-obvious, briefly call it out so the author knows it's intentional.

assets/skills/delegation-protocol/SKILL.md

@@ -0,0 +1,69 @@
+---
+description: Structured 6-section delegation template and session-continuity rules for orchestrating sub-agents. Load before spawning any agent.
+---
+You are delegating work to a sub-agent. The sub-agent has not seen the codebase or the conversation — your prompt IS its entire context. Treat delegation as writing a contract: explicit, scoped, and verifiable.
+
+## The 6-section template (every delegation)
+
+Every `agent__spawn` prompt MUST include all six sections. Vague prompts produce vague results and waste tokens on re-exploration the orchestrator already did.
+
+```
+## TASK
+[One atomic goal. One verb. One outcome. No "and also".]
+
+## EXPECTED OUTCOME
+[Concrete deliverables and success criteria. "I will know this is done when ..."]
+
+## REQUIRED TOOLS
+[Explicit allowlist: fs_read, fs_grep, etc. Prevents tool sprawl.]
+
+## MUST DO
+[Exhaustive requirements. Leave nothing implicit. If you'd be annoyed by the agent not doing X, list X.]
+
+## MUST NOT DO
+[Forbidden actions. Anticipate rogue behavior. "Do not modify files outside src/auth/."]
+
+## CONTEXT
+[File paths, code snippets, existing patterns, constraints. Paste actual code lines from prior exploration — not just file paths.]
+```
+
+## Session continuity (NON-NEGOTIABLE)
+
+Every `agent__spawn` result includes a session_id. **Use it.**
+
+- Task failed/incomplete → resume with `session_id` + a tight "Fix: <error>" prompt.
+- Follow-up on a result → resume with `session_id` + "Also: <question>".
+- Multi-turn with the same agent → always resume. Never start fresh.
+
+Starting a fresh agent for a follow-up forces it to re-read every file it already read. That's 70%+ wasted tokens, plus the agent loses the reasoning it built up.
+
+After every delegation, **store the session_id compression-safe** for potential continuation. Long sessions compress: chat history gets replaced by a summary, and a session_id that exists only in chat history is unresumable afterward. Embed it in the todo item for that work — `todo__add "Implement auth endpoint (coder ses_abc123)"` — or in your run-state memory file. The todo list and memory survive compression; the conversation does not.
+
+## Skill nudges to delegates
+
+Sub-agents have their own skills. Nudge them in the CONTEXT section:
+
+> "Load `code-review` before evaluating the diff."
+> "Load `frontend-ui-ux` before editing component files."
+> "Load `git-master` before touching history."
+
+A one-line nudge saves the delegate a `skill__list` turn.
+
+## Verification after delegation
+
+A delegation is NOT complete when the sub-agent returns. It is complete when YOU have verified:
+
+1. Did it work as expected? (Did the file change? Did the test pass?)
+2. Did it follow existing codebase patterns?
+3. Did the EXPECTED OUTCOME actually materialize?
+4. Did it respect MUST DO and MUST NOT DO?
+
+If any answer is no → resume the session with a corrective prompt. Do not re-spawn from scratch.
+
+## Anti-patterns
+
+- "Follow existing patterns" with no snippet → agent guesses, often wrong
+- Multi-goal prompts → agent does the easy one, skips the rest
+- Missing MUST NOT DO → agent over-reaches into unrelated files
+- Discarding session_id on failure → forced re-exploration, wasted tokens
+- Re-spawning instead of resuming for a 1-line fix → 10x cost

assets/skills/design-session/SKILL.md

@@ -0,0 +1,79 @@
+---
+description: AI-first design decomposition for any project. Given a design doc or topic, ground in the actual codebase, produce (or refine) a PLAN file with problem, approach, alternatives, constraints, and a task breakdown sized to ~1 engineer-day per task with measurable acceptance criteria. The plan is written to be a self-contained "sealed container" for context-free implementers. Grants filesystem access for grounding and for writing the plan.
+enabled_tools: fs_read, fs_grep, fs_glob, fs_ls, fs_cat, fs_write
+---
+You are decomposing a design doc (or topic) into an executable plan. The output is ONE plan file plus a task breakdown that context-free LLM implementers will execute later with zero access to this conversation. Everything they need must be on the page or pointed to — see the "sealed container" standard below.
+
+## Inputs
+
+- A design doc (path or pasted), or a one-line problem statement.
+- The target project directory (ground truth for all claims).
+- The plans directory where the PLAN file lands.
+
+## Step 1 — Ground before proposing
+
+Plans written from memory rot on contact with the code. Before writing anything:
+
+- Read the project's own orientation docs (`CLAUDE.md`, `AGENTS.md`, `CONTRIBUTING.md`, `README.md` at the project root) — conventions constrain the design.
+- Read the code the design touches: entry points, the modules to be changed, neighboring examples of the patterns to follow, existing tests.
+- `fs_grep` every symbol the design doc references — confirm it exists and is spelled right. Note explicitly: what already exists, what would be added, what would change.
+- Verify build/test commands actually exist (`Makefile`, `justfile`, `package.json` scripts, CI config).
+
+## Step 2 — The proposal
+
+Produce a structured proposal (iterate with the user when interactive; in autonomous runs, resolve what the doc + code answer and flag the rest as open questions):
+
+- **Problem** — one paragraph; state assumptions explicitly.
+- **Scope** — In / Out. Call out tempting adjacent work being deferred.
+- **Approach** — concrete: name files, symbols, data flow, migrations. Reference existing patterns by path.
+- **Alternatives considered** — table of alternative → why rejected. Settled decisions carry their one-line reason (an unrecorded decision WILL be re-litigated by an implementer).
+- **Constraints and risks** — conventions the design must respect; ordering dependencies; things you're uncertain about, flagged clearly.
+- **Open questions** — ONLY questions the codebase cannot answer (business rules, priority calls). If none, say "No open questions."
+- **Task breakdown** — see below.
+
+## Task breakdown rules
+
+| Rule | Why |
+|---|---|
+| **One task ≈ one engineer-day** | Variable task sizes destroy progress signal; anything larger gets decomposed NOW, not mid-run |
+| Each task independently implementable and verifiable | It builds and its tests pass without later tasks existing |
+| Explicit, acyclic dependencies (`blocked_by`) | Execution order must be derivable from the breakdown alone |
+| Each task states WHERE (files/packages) and WHAT (observable outcome) | "Implement service layer" is not a task; "internal/foo/service.go: add Create/Get with validation — returns 400 on missing name" is |
+| Measurable acceptance criteria per task | Criteria become the tests; "works correctly" is unmeasurable |
+| Flag ⚠️ low-confidence sizing with the reason | Honest sizing beats optimistic sizing |
+
+## Step 3 — Write the PLAN file
+
+Write `PLAN-<slug>.md` (kebab-case slug from the topic; verify no collision) to the plans directory:
+
+```markdown
+---
+slug: <slug>
+status: draft        # draft | active | implemented
+created: YYYY-MM-DD
+---
+
+# <Title>
+
+## Problem
+## Scope (In / Out)
+## Approach
+## Alternatives considered
+## Constraints and risks
+## Open questions
+## Task breakdown
+
+| # | Task | Size | blocked_by | Notes |
+|---|------|------|-----------|-------|
+```
+
+The plan is the implementers' entire context. Write for the "sealed container" standard: every question an implementer will hit is either answered inline or delegated via a pointer to the exact file/doc that answers it (where infra code goes, what DB tech, which layout to mirror, exact test commands). Paste short code snippets for load-bearing patterns — a path alone forces re-exploration; a stale claim fails the executor mid-implementation.
+
+## Anti-patterns
+
+- Proposing before reading the code — a design ungrounded in the actual codebase is fiction.
+- "As discussed" / "per our conversation" — the implementer has no conversation.
+- Tasks larger than a day hiding an "and then also…".
+- Acceptance criteria describing implementation ("uses a for loop") instead of behavior.
+- Open questions the code could have answered — grep first, ask last.
+- Unrecorded decisions — every settled fork carries its reason.

assets/skills/diagnostics/SKILL.md

@@ -0,0 +1,44 @@
+---
+description: Systematic troubleshooting of technical issues (services, networking, containers, OS) by running diagnostic commands directly instead of asking the user to.
+enabled_tools: execute_command
+---
+A technical problem needs diagnosing. Apply this methodology strictly. Use the `execute_command` tool to gather
+evidence yourself — never ask the user to run commands and paste output back.
+
+## Loop
+
+1. **Reproduce first.** Run the failing thing and read the actual error before theorizing.
+2. **Ask "what changed?"** Updates, config edits, reboots, expirations. Check recent history early.
+3. **Cheap checks first.** Service running/enabled? Interface up? Disk full? DNS resolving? Clock right?
+4. **Isolate by layer, one variable at a time.** Network: link → IP → routing → DNS → transport → app.
+   Software: process → logs → config → deps/permissions → environment. Containers: daemon → image → container →
+   logs → mounts/networks → host.
+5. **State each hypothesis in one line before testing it.** Pivot openly when disproved.
+6. **Fix root cause, then verify** by re-running the original failing operation. No verification, no fix.
+
+## When to Stop Gathering Evidence
+
+Once you have two or more independent pieces of evidence pointing to the same root cause, **stop gathering and deliver your diagnosis**. Do not add more verification steps to verify your verification. If you notice yourself thinking "let me just confirm one more thing" after you have already reached a conclusion, that is the signal to stop and explain the diagnosis instead. More data is not always better — a timely diagnosis with strong evidence beats an exhaustive audit.
+
+## Command Discipline
+
+- Non-interactive and bounded, always: `--no-pager`, `-n`/`--since` on logs, `timeout 10` on anything that might
+  hang, `-c` on ping. No TUIs — use batch modes.
+- Unprivileged first; `sudo` only when required, stating why.
+- Web-search exact quoted error strings (with software name + version) for unfamiliar errors.
+
+## Safety Tiers
+
+1. **Read-only** (status, logs, ls, cat, ping, dig): run freely.
+2. **Reversible changes** (service restart, interface bounce, config edit): announce in one sentence, back up files
+   first (`cp file file.bak.$(date +%s)`), then do it.
+3. **Destructive** (data/volume deletion, formatting, `dd`, package removal, firewall flush): require explicit user
+   confirmation with the exact command and a rollback plan. Never on your own judgment.
+
+Redact any secrets appearing in command output. Never disable security controls as a "fix". Stop and present options
+if evidence suggests failing hardware or data-loss risk.
+
+## Reporting
+
+Lead with findings, show trimmed key evidence, and close resolved issues with: root cause → fix → verification →
+prevention.