|
|
d822180205
|
fix: updated the error for unauthenticated oauth to include the REPL .authenticated command
|
2026-03-28 11:57:01 -06:00 |
|
|
|
d78820dcd4
|
fix: Claude code system prompt injected into claude requests to make them valid once again
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-03-17 10:44:50 -06:00 |
|
|
|
d43c4232a2
|
fix: Do not inject tools when models don't support them; detect this conflict before API calls happen
|
2026-03-17 09:35:51 -06:00 |
|
|
|
f41c85b703
|
style: Applied formatting across new inquire files
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-03-16 12:39:20 -06:00 |
|
|
|
9e056bdcf0
|
feat: Added support for specifying a custom response to multiple-choice prompts when nothing suits the user's needs
|
2026-03-16 12:37:47 -06:00 |
|
|
|
d6022b9f98
|
feat: Supported theming in the inquire prompts in the REPL
|
2026-03-16 12:36:20 -06:00 |
|
|
|
b2dbdfb4b1
|
feat: Support for Gemini OAuth
|
2026-03-12 13:29:47 -06:00 |
|
|
|
063e198f96
|
refactor: Made the oauth module more generic so it can support loopback OAuth (not just manual)
|
2026-03-12 13:28:09 -06:00 |
|
|
|
eb4d1c02f4
|
feat: Support authenticating or refreshing OAuth for supported clients from within the REPL
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-03-11 13:07:27 -06:00 |
|
|
|
c428990900
|
fix: the updated regex for secrets injection broke MCP server secrets interpolation because the regex greedily matched on new lines, replacing too much content. This fix just ignores commented out lines in YAML files by skipping commented out lines.
|
2026-03-11 12:55:28 -06:00 |
|
|
|
03b9cc70b9
|
feat: Allow first-runs to select OAuth for supported providers
|
2026-03-11 12:01:17 -06:00 |
|
|
|
3fa0eb832c
|
fix: Don't try to inject secrets into commented-out lines in the config
|
2026-03-11 11:11:09 -06:00 |
|
|
|
83f66e1061
|
feat: Support OAuth authentication flows for Claude
|
2026-03-11 11:10:48 -06:00 |
|
|
|
101a9cdd6e
|
style: Applied formatting to MCP module
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-02-20 15:28:21 -07:00 |
|
|
|
e6e99b6926
|
feat: Improved MCP server spinup and spindown when switching contexts or settings in the REPL: Modify existing config rather than stopping all servers always and re-initializing if unnecessary
|
2026-02-20 14:36:34 -07:00 |
|
|
|
15a293204f
|
fix: Improved sub-agent stdout and stderr output for users to follow
|
2026-02-20 13:47:28 -07:00 |
|
|
|
60493728a0
|
fix: Inject agent variables into environment variables for global tool calls when invoked from agents to modify global tool behavior
|
2026-02-20 11:38:24 -07:00 |
|
|
|
32a55652fe
|
fix: Fixed a bug in the new supervisor and todo built-ins that was causing errors with OpenAI models
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-02-18 14:52:57 -07:00 |
|
|
|
2bb8cf5f73
|
fix: Removed the now unnecessary is_stale_response that was breaking auto-continuing with parallel agents
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-02-18 13:36:25 -07:00 |
|
|
|
fbac446859
|
style: Applied formatting to the function module
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-02-18 13:20:18 -07:00 |
|
|
|
f91cf2e346
|
build: Upgraded to the most recent version of rmcp
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-02-18 12:28:52 -07:00 |
|
|
|
c1902a69d1
|
feat: Created a CodeRabbit-style code-reviewer agent
|
2026-02-18 12:16:59 -07:00 |
|
|
|
655ee2a599
|
fix: Bypassed enabled_tools for user interaction tools so if function calling is enabled at all, the LLM has access to the user interaction tools when in REPL mode
|
2026-02-18 11:25:25 -07:00 |
|
|
|
128a8f9a9c
|
feat: Added configuration option in agents to indicate the timeout for user input before proceeding (defaults to 5 minutes)
|
2026-02-18 11:24:47 -07:00 |
|
|
|
b1be9443e7
|
feat: Added support for sub-agents to escalate user interaction requests from any depth to the parent agents for user interactions
|
2026-02-18 11:06:15 -07:00 |
|
|
|
7b12c69ebf
|
feat: 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
|
2026-02-18 11:05:43 -07:00 |
|
|
|
69ad584137
|
fix: When parallel agents run, only write to stdout from the parent and only display the parent's throbber
|
2026-02-18 09:59:24 -07:00 |
|
|
|
313058e70a
|
refactor: Cleaned up some left-over implementation stubs
|
2026-02-18 09:13:39 -07:00 |
|
|
|
ea96d9ba3d
|
fix: Forgot to implement support for failing a task and keep all dependents blocked
|
2026-02-18 09:13:11 -07:00 |
|
|
|
7884adc7c1
|
fix: Clean up orphaned sub-agents when the parent agent
|
2026-02-18 09:12:32 -07:00 |
|
|
|
5e9c31595e
|
fix: Forgot to automatically add the bidirectional communication back up to parent agents from sub-agents (i.e. need to be able to check inbox and send messages)
|
2026-02-17 16:11:35 -07:00 |
|
|
|
39d9b25e47
|
feat: Added an agent configuration property that allows auto-injecting sub-agent spawning instructions (when using the built-in sub-agent spawning system)
|
2026-02-17 15:49:40 -07:00 |
|
|
|
b86f76ddb9
|
feat: Auto-dispatch support of sub-agents and support for the teammate pattern between subagents
|
2026-02-17 15:18:27 -07:00 |
|
|
|
7f267a10a1
|
docs: Initial documentation cleanup of parallel agent MVP
|
2026-02-17 14:30:28 -07:00 |
|
|
|
cdafdff281
|
fix: Agent delegation tools were not being passed into the {{__tools__}} placeholder so agents weren't delegating to subagents
|
2026-02-17 14:19:22 -07:00 |
|
|
|
60ad83d6d9
|
feat: Full passive task queue integration for parallelization of subagents
|
2026-02-17 13:42:53 -07:00 |
|
|
|
44c03ccf4f
|
feat: Implemented initial scaffolding for built-in sub-agent spawning tool call operations
|
2026-02-17 11:48:31 -07:00 |
|
|
|
af933bbb29
|
feat: Initial models for agent parallelization
|
2026-02-17 11:27:55 -07:00 |
|
|
|
d2f8f995f0
|
feat: 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
CI / All (macos-latest) (push) Has been cancelled
CI / All (ubuntu-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-02-13 17:45:56 -07:00 |
|
|
|
d2e541c5c0
|
docs: Updated todo-system docs
|
2026-02-13 15:13:37 -07:00 |
|
|
|
b11797ea1c
|
fix: Improved continuation prompt to not make broad todo-items
|
2026-02-09 15:36:57 -07:00 |
|
|
|
70c2d411ae
|
fix: Allow auto-continuation to work in agents after a session is compressed and if there's still unfinish items in the to-do list
|
2026-02-09 15:21:39 -07:00 |
|
|
|
a935add2a7
|
feat: Implemented a built-in task management system to help smaller LLMs complete larger multistep tasks and minimize context drift
|
2026-02-09 12:49:06 -07:00 |
|
|
|
8a37a88ffd
|
feat: Improved tool and MCP invocation error handling by returning stderr to the model when it is available
|
2026-02-04 12:00:21 -07:00 |
|
|
|
8f66cac680
|
feat: Added variable interpolation for conversation starters in agents
CI / All (ubuntu-latest) (push) Failing after 5m27s
CI / All (macos-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-02-04 10:51:59 -07:00 |
|
|
|
d5e0728532
|
feat: Implemented retry logic for failed tool invocations so the LLM can learn from the result and try again; Also implemented chain loop detection to prevent loops
CI / All (ubuntu-latest) (push) Failing after 5m27s
CI / All (macos-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
|
2026-02-01 17:06:16 -07:00 |
|
|
|
25c0885dcc
|
fix: Claude tool calls work incorrectly when tool doesn't require any arguments or flags; would provide an empty JSON object or error on no args
|
2026-02-01 17:05:36 -07:00 |
|
|
|
f56ed7d005
|
feat: Added gemini-3-pro to the supported vertexai models
|
2026-01-30 19:03:41 -07:00 |
|
|
|
d79e4b9dff
|
Fixed some typos in tool call error messages
|
2026-01-30 12:25:57 -07:00 |
|
|
|
f2de196e22
|
fix: Fixed a bug where --agent-variable values were not being passed to the agents
|
2026-01-23 14:15:59 -07:00 |
|
