Compare commits

...
Author SHA1 Message Date
Dark-Alex-17 ae96a7e031 fmt: applied formatting 2026-08-12 19:35:27 -06:00
Dark-Alex-17 2658ca776e feat: Installed duckdb into the coyote image
CI / All (ubuntu-latest) (push) Failing after 29s
CI / All (macos-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
2026-08-12 19:29:26 -06:00
Dark-Alex-17 f8682102a0 docs: Added duckdb prerequisite
CI / All (ubuntu-latest) (push) Failing after 29s
CI / All (macos-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
2026-08-12 19:07:31 -06:00
Dark-Alex-17 3fa0f5c428 feat: Support managing MCP servers from the CLI directly
CI / All (ubuntu-latest) (push) Failing after 29s
CI / All (macos-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
2026-08-12 17:55:33 -06:00
Dark-Alex-17 68135b97d1 feat: append new built-in rag__query function to RAG contexts to allow further querying by LLMs
CI / All (ubuntu-latest) (push) Failing after 29s
CI / All (macos-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
2026-08-12 16:25:50 -06:00
Dark-Alex-17 b87a3460c4 fix: detect duplicate tool call IDs client-side before sending to Claude
CI / All (ubuntu-latest) (push) Failing after 29s
CI / All (macos-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
2026-08-12 13:33:24 -06:00
Dark-Alex-17 cb23da6490 feat: support static file bundling with sbx-mixins
CI / All (ubuntu-latest) (push) Failing after 29s
CI / All (macos-latest) (push) Has been cancelled
CI / All (windows-latest) (push) Has been cancelled
2026-08-12 13:18:38 -06:00
14 changed files with 1228 additions and 19 deletions
+15
View File
@@ -36,6 +36,21 @@ RUN set -euo pipefail; \
install -m 0755 "$TMPDIR/usql_static" /usr/local/bin/usql; \ install -m 0755 "$TMPDIR/usql_static" /usr/local/bin/usql; \
rm -rf "$TMPDIR" rm -rf "$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 USER 1000
RUN curl -LsSf https://astral.sh/uv/install.sh | sh && \ RUN curl -LsSf https://astral.sh/uv/install.sh | sh && \
+2
View File
@@ -70,6 +70,8 @@ Coyote requires the following tools to be installed on your system:
* **Cargo:** `cargo install ast-grep --locked` * **Cargo:** `cargo install ast-grep --locked`
* **npm:** `npm i -g @ast-grep/cli` * **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` * 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 Coyote, 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. etc., and they are used within agents and tools.
+161 -1
View File
@@ -13,6 +13,20 @@ use is_terminal::IsTerminal;
use std::collections::HashSet; use std::collections::HashSet;
use std::io::{Read, stdin}; use std::io::{Read, stdin};
#[derive(clap::ValueEnum, Debug, Clone, Copy, PartialEq, Eq)]
pub enum McpTransportArg {
Stdio,
Http,
Sse,
}
#[derive(clap::ValueEnum, Debug, Clone, Copy, PartialEq, Eq, Default)]
pub enum McpScopeArg {
#[default]
User,
Workspace,
}
#[derive(Parser, Debug)] #[derive(Parser, Debug)]
#[command(author, version, about, long_about = None)] #[command(author, version, about, long_about = None)]
#[command( #[command(
@@ -41,10 +55,15 @@ use std::io::{Read, stdin};
"list_skills", "skill", "tail_logs", "completions", "update", "list_skills", "skill", "tail_logs", "completions", "update",
]) ])
), ),
group(
ArgGroup::new("mcp-action")
.args(["mcp_add", "mcp_remove", "mcp_list", "mcp_get"])
.multiple(false)
),
)] )]
pub struct Cli { pub struct Cli {
/// Input text /// Input text
#[arg(trailing_var_arg = true)] #[arg(allow_hyphen_values = true)]
text: Vec<String>, text: Vec<String>,
/// Select a LLM model /// Select a LLM model
@@ -224,6 +243,57 @@ pub struct Cli {
#[arg(long, exclusive = true, value_name = "SERVER_NAME", help_heading = "Authentication", add = ArgValueCompleter::new(mcp_server_completer))] #[arg(long, exclusive = true, value_name = "SERVER_NAME", help_heading = "Authentication", add = ArgValueCompleter::new(mcp_server_completer))]
pub auth_mcp: Option<String>, pub auth_mcp: Option<String>,
/// Add an MCP server. Use `-- <cmd> [args...]` for stdio, or `--url <URL>` for http/sse.
#[arg(long, value_name = "NAME", help_heading = "MCP Servers")]
pub mcp_add: Option<String>,
/// Remove an MCP server by name
#[arg(long, value_name = "NAME", help_heading = "MCP Servers", add = ArgValueCompleter::new(mcp_server_completer))]
pub mcp_remove: Option<String>,
/// List all configured MCP servers (user + workspace scopes)
#[arg(long, help_heading = "MCP Servers")]
pub mcp_list: bool,
/// Show the JSON config for one MCP server
#[arg(long, value_name = "NAME", help_heading = "MCP Servers", add = ArgValueCompleter::new(mcp_server_completer))]
pub mcp_get: Option<String>,
/// Transport for --mcp-add: stdio (default when `--` present), http, or sse
#[arg(
long,
value_enum,
value_name = "TRANSPORT",
help_heading = "MCP Servers"
)]
pub transport: Option<McpTransportArg>,
/// URL for http/sse MCP server (used with --mcp-add)
#[arg(long, value_name = "URL", help_heading = "MCP Servers")]
pub url: Option<String>,
/// Scope for MCP config: user (~/.config/coyote/functions/mcp.json) or workspace (./.coyote/mcp.json). Default: user
#[arg(long, value_enum, value_name = "SCOPE", help_heading = "MCP Servers")]
pub scope: Option<McpScopeArg>,
/// Environment variable for stdio MCP server (repeatable): --env KEY=VALUE
#[arg(long, value_name = "KEY=VALUE", help_heading = "MCP Servers")]
pub env: Vec<String>,
/// HTTP header for http/sse MCP server (repeatable): --header "Name: Value"
#[arg(long, value_name = "HEADER", help_heading = "MCP Servers")]
pub header: Vec<String>,
/// Working directory for stdio MCP server
#[arg(long, value_name = "PATH", value_hint = ValueHint::AnyPath, help_heading = "MCP Servers")]
pub cwd: Option<String>,
/// OAuth client ID for http/sse MCP server
#[arg(long, value_name = "ID", help_heading = "MCP Servers")]
pub client_id: Option<String>,
/// OAuth client secret for http/sse MCP server (use {{NAME}} to reference a vault secret)
#[arg(long, value_name = "SECRET", help_heading = "MCP Servers")]
pub client_secret: Option<String>,
/// OAuth callback port for http/sse MCP server
#[arg(long, value_name = "PORT", help_heading = "MCP Servers")]
pub callback_port: Option<u16>,
/// OAuth redirect host for http/sse MCP server
#[arg(long, value_name = "HOST", help_heading = "MCP Servers")]
pub redirect_host: Option<String>,
/// Overwrite an existing MCP server (with --mcp-add) or skip confirmation (with --mcp-remove)
#[arg(long, help_heading = "MCP Servers")]
pub mcp_force: bool,
/// Launch Coyote inside a Docker sandbox (via `sbx`); name defaults to current directory basename /// Launch Coyote inside a Docker sandbox (via `sbx`); name defaults to current directory basename
#[arg(long, value_name = "NAME", help_heading = "Sandbox")] #[arg(long, value_name = "NAME", help_heading = "Sandbox")]
pub sandbox: Option<Option<String>>, pub sandbox: Option<Option<String>>,
@@ -254,6 +324,15 @@ pub struct Cli {
/// Generate static shell completion scripts /// Generate static shell completion scripts
#[arg(long, value_name = "SHELL", value_enum, help_heading = "Shell")] #[arg(long, value_name = "SHELL", value_enum, help_heading = "Shell")]
pub completions: Option<ShellCompletion>, pub completions: Option<ShellCompletion>,
/// Stdio command for --mcp-add: everything after `--` is passed to the server verbatim
#[arg(
last = true,
allow_hyphen_values = true,
value_name = "CMD",
help_heading = "MCP Servers"
)]
pub mcp_command: Vec<String>,
} }
impl Cli { impl Cli {
@@ -633,4 +712,85 @@ mod tests {
fn parse_sandbox_is_exclusive() { fn parse_sandbox_is_exclusive() {
assert!(Cli::try_parse_from(["coyote", "--sandbox", "--agent", "foo"]).is_err()); assert!(Cli::try_parse_from(["coyote", "--sandbox", "--agent", "foo"]).is_err());
} }
#[test]
fn parse_mcp_add_stdio_with_trailing_command() {
let cli = parse(&[
"--mcp-add",
"myserver",
"--",
"npx",
"some-server",
"--flag",
"arg1",
]);
assert_eq!(cli.mcp_add, Some("myserver".to_string()));
assert_eq!(
cli.mcp_command,
vec!["npx", "some-server", "--flag", "arg1"]
);
assert!(cli.text.is_empty());
}
#[test]
fn parse_mcp_add_stdio_with_env_and_command() {
let cli = parse(&[
"--mcp-add",
"s",
"--env",
"API_KEY={{API_KEY}}",
"--env",
"MODE=dev",
"--",
"npx",
"srv",
]);
assert_eq!(cli.mcp_add, Some("s".to_string()));
assert_eq!(cli.env, vec!["API_KEY={{API_KEY}}", "MODE=dev"]);
assert_eq!(cli.mcp_command, vec!["npx", "srv"]);
}
#[test]
fn parse_mcp_add_http_with_header() {
let cli = parse(&[
"--mcp-add",
"notion",
"--transport",
"http",
"--url",
"https://mcp.notion.com/mcp",
"--header",
"Authorization: Bearer {{NOTION_TOKEN}}",
]);
assert_eq!(cli.mcp_add, Some("notion".to_string()));
assert!(matches!(cli.transport, Some(McpTransportArg::Http)));
assert_eq!(cli.url, Some("https://mcp.notion.com/mcp".to_string()));
assert_eq!(cli.header, vec!["Authorization: Bearer {{NOTION_TOKEN}}"]);
assert!(cli.mcp_command.is_empty());
}
#[test]
fn parse_mcp_list_flag() {
let cli = parse(&["--mcp-list"]);
assert!(cli.mcp_list);
}
#[test]
fn parse_mcp_scope_workspace() {
let cli = parse(&["--mcp-list", "--scope", "workspace"]);
assert!(cli.mcp_list);
assert!(matches!(cli.scope, Some(McpScopeArg::Workspace)));
}
#[test]
fn parse_mcp_action_group_is_exclusive() {
assert!(Cli::try_parse_from(["coyote", "--mcp-list", "--mcp-get", "foo"]).is_err());
}
#[test]
fn parse_trailing_text_unchanged_without_dash_dash() {
let cli = parse(&["hello", "world"]);
assert_eq!(cli.text, vec!["hello", "world"]);
assert!(cli.mcp_command.is_empty());
}
} }
+115 -5
View File
@@ -1,3 +1,4 @@
use std::collections::HashSet;
use std::mem; use std::mem;
use super::access_token::get_access_token; use super::access_token::get_access_token;
@@ -368,17 +369,32 @@ pub fn claude_build_chat_completions_body(
] ]
} else { } else {
// One pair per round: Claude can reuse tool_use IDs across API calls. // One pair per round: Claude can reuse tool_use IDs across API calls.
// A round boundary is detected by the presence of round text, but
// rounds where the model emitted only tool calls (no narration)
// carry no text marker. As a backstop, also split whenever a
// tool_use ID would repeat within the current assistant message —
// the API rejects duplicate tool_use IDs in a single message.
let mut messages = vec![]; let mut messages = vec![];
let mut assistant_parts: Vec<serde_json::Value> = vec![]; let mut assistant_parts: Vec<serde_json::Value> = vec![];
let mut user_parts: Vec<serde_json::Value> = vec![]; let mut user_parts: Vec<serde_json::Value> = vec![];
let mut chunk_ids: HashSet<&str> = HashSet::new();
for (index, tool_result) in tool_results.iter().enumerate() { for (index, tool_result) in tool_results.iter().enumerate() {
if index > 0 && tool_result.text.is_some() { let id_collision = tool_result
.call
.id
.as_deref()
.is_some_and(|id| chunk_ids.contains(id));
if index > 0 && (tool_result.text.is_some() || id_collision) {
messages.push( messages.push(
json!({ "role": "assistant", "content": assistant_parts }), json!({ "role": "assistant", "content": assistant_parts }),
); );
messages.push(json!({ "role": "user", "content": user_parts })); messages.push(json!({ "role": "user", "content": user_parts }));
assistant_parts = vec![]; assistant_parts = vec![];
user_parts = vec![]; user_parts = vec![];
chunk_ids.clear();
}
if let Some(id) = tool_result.call.id.as_deref() {
chunk_ids.insert(id);
} }
for block in &tool_result.thinking { for block in &tool_result.thinking {
assistant_parts.push(json!(block)); assistant_parts.push(json!(block));
@@ -485,10 +501,7 @@ pub fn claude_extract_chat_completions(data: &Value) -> Result<ChatCompletionsOu
if let Some(v) = item["thinking"].as_str() { if let Some(v) = item["thinking"].as_str() {
thinking.push(ThinkingBlock::Thinking { thinking.push(ThinkingBlock::Thinking {
thinking: v.to_string(), thinking: v.to_string(),
signature: item["signature"] signature: item["signature"].as_str().unwrap_or_default().to_string(),
.as_str()
.unwrap_or_default()
.to_string(),
}); });
} }
} }
@@ -535,3 +548,100 @@ pub fn claude_extract_chat_completions(data: &Value) -> Result<ChatCompletionsOu
}; };
Ok(output) Ok(output)
} }
#[cfg(test)]
mod tests {
use super::*;
use crate::function::{ToolCall, ToolResult};
fn tool_result(id: &str, text: Option<&str>) -> ToolResult {
ToolResult {
call: ToolCall::new("fs_read".into(), json!({"path": "x"}), Some(id.into())),
output: json!("ok"),
text: text.map(|t| t.to_string()),
thinking: vec![],
}
}
fn build_body(tool_results: Vec<ToolResult>) -> Value {
let data = ChatCompletionsData {
messages: vec![
Message::new(MessageRole::User, MessageContent::Text("hello".to_string())),
Message::new(
MessageRole::Assistant,
MessageContent::ToolCalls(MessageContentToolCalls {
tool_results,
text: String::new(),
sequence: true,
}),
),
],
temperature: None,
top_p: None,
reasoning_effort: None,
functions: None,
stream: false,
};
claude_build_chat_completions_body(data, &Model::new("claude", "claude-test")).unwrap()
}
fn assert_unique_tool_use_ids_per_message(body: &Value) {
for message in body["messages"].as_array().unwrap() {
let Some(content) = message["content"].as_array() else {
continue;
};
let mut seen = HashSet::new();
for block in content {
if block["type"] == "tool_use" {
let id = block["id"].as_str().unwrap();
assert!(
seen.insert(id.to_string()),
"duplicate tool_use id `{id}` within a single assistant message: {message}"
);
}
}
}
}
#[test]
fn sequence_splits_on_round_text() {
let body = build_body(vec![
tool_result("toolu_A", None),
tool_result("toolu_B", None),
tool_result("toolu_C", Some("running another tool")),
]);
let messages = body["messages"].as_array().unwrap();
assert_eq!(messages.len(), 5, "body: {body}");
assert_unique_tool_use_ids_per_message(&body);
}
#[test]
fn sequence_splits_on_reused_id_in_textless_round() {
let body = build_body(vec![
tool_result("toolu_A", None),
tool_result("toolu_B", None),
tool_result("toolu_A", None),
]);
let messages = body["messages"].as_array().unwrap();
assert_eq!(messages.len(), 5, "body: {body}");
assert_unique_tool_use_ids_per_message(&body);
}
#[test]
fn sequence_keeps_textless_rounds_merged_when_ids_are_unique() {
let body = build_body(vec![
tool_result("toolu_A", None),
tool_result("toolu_B", None),
tool_result("toolu_C", None),
]);
let messages = body["messages"].as_array().unwrap();
assert_eq!(messages.len(), 3, "body: {body}");
assert_unique_tool_use_ids_per_message(&body);
}
}
+4
View File
@@ -248,6 +248,10 @@ impl Agent {
} }
} }
if rag.is_some() && app.function_calling_support && graph_for_rag.is_none() {
functions.append_rag_query_functions();
}
agent_config.replace_tools_placeholder(&functions); agent_config.replace_tools_placeholder(&functions);
Ok(Self { Ok(Self {
+18 -5
View File
@@ -16,8 +16,9 @@ use super::{MessageContentToolCalls, prompts};
use crate::client::{Model, ModelType, list_models}; use crate::client::{Model, ModelType, list_models};
use crate::function::{ use crate::function::{
FunctionDeclaration, Functions, ToolCallTracker, ToolResult, memory::MEMORY_FUNCTION_PREFIX, FunctionDeclaration, Functions, ToolCallTracker, ToolResult, memory::MEMORY_FUNCTION_PREFIX,
skill::SKILL_FUNCTION_PREFIX, supervisor::SUPERVISOR_FUNCTION_PREFIX, rag_query::RAG_FUNCTION_PREFIX, skill::SKILL_FUNCTION_PREFIX,
todo::TODO_FUNCTION_PREFIX, user_interaction::USER_FUNCTION_PREFIX, supervisor::SUPERVISOR_FUNCTION_PREFIX, todo::TODO_FUNCTION_PREFIX,
user_interaction::USER_FUNCTION_PREFIX,
}; };
use crate::mcp::{ use crate::mcp::{
MCP_DESCRIBE_META_FUNCTION_NAME_PREFIX, MCP_INVOKE_META_FUNCTION_NAME_PREFIX, MCP_DESCRIBE_META_FUNCTION_NAME_PREFIX, MCP_INVOKE_META_FUNCTION_NAME_PREFIX,
@@ -715,6 +716,7 @@ impl RequestContext {
pub fn exit_rag(&mut self) -> Result<()> { pub fn exit_rag(&mut self) -> Result<()> {
self.rag.take(); self.rag.take();
self.tool_scope.functions.remove_rag_query_functions();
Ok(()) Ok(())
} }
@@ -1137,6 +1139,7 @@ impl RequestContext {
&& !v.name.starts_with("agent__") && !v.name.starts_with("agent__")
&& !v.name.starts_with("memory__") && !v.name.starts_with("memory__")
&& !v.name.starts_with("skill__") && !v.name.starts_with("skill__")
&& !v.name.starts_with("rag__")
}) })
.map(|v| v.name.clone()) .map(|v| v.name.clone())
.collect() .collect()
@@ -1957,7 +1960,8 @@ impl RequestContext {
|| (!matches!(role.skills_enabled(), Some(false)) || (!matches!(role.skills_enabled(), Some(false))
&& v.name.starts_with(SKILL_FUNCTION_PREFIX)) && v.name.starts_with(SKILL_FUNCTION_PREFIX))
|| (self.auto_continue_config().enabled || (self.auto_continue_config().enabled
&& v.name.starts_with(TODO_FUNCTION_PREFIX))) && v.name.starts_with(TODO_FUNCTION_PREFIX))
|| v.name.starts_with(RAG_FUNCTION_PREFIX))
&& !existing.contains(&v.name) && !existing.contains(&v.name)
}) })
.cloned() .cloned()
@@ -1987,6 +1991,7 @@ impl RequestContext {
|| v.name.starts_with(TODO_FUNCTION_PREFIX) || v.name.starts_with(TODO_FUNCTION_PREFIX)
|| v.name.starts_with(SUPERVISOR_FUNCTION_PREFIX) || v.name.starts_with(SUPERVISOR_FUNCTION_PREFIX)
|| v.name.starts_with(MEMORY_FUNCTION_PREFIX) || v.name.starts_with(MEMORY_FUNCTION_PREFIX)
|| v.name.starts_with(RAG_FUNCTION_PREFIX)
}); });
} }
@@ -3467,6 +3472,12 @@ impl RequestContext {
if self.should_register_memory_tools() { if self.should_register_memory_tools() {
functions.append_memory_functions(); functions.append_memory_functions();
} }
if self.rag.is_some()
&& app.function_calling_support
&& !self.agent.as_ref().is_some_and(|a| a.is_graph())
{
functions.append_rag_query_functions();
}
let tool_tracker = self.tool_scope.tool_tracker.clone(); let tool_tracker = self.tool_scope.tool_tracker.clone();
self.tool_scope = ToolScope { self.tool_scope = ToolScope {
@@ -4136,7 +4147,7 @@ impl RequestContext {
super::TEMP_RAG_NAME, super::TEMP_RAG_NAME,
&rag_path, &rag_path,
&[], &[],
abort_signal, abort_signal.clone(),
false, false,
) )
.await?, .await?,
@@ -4172,10 +4183,11 @@ impl RequestContext {
}; };
self.rag = Some(rag); self.rag = Some(rag);
self.rag_key = rag_key; self.rag_key = rag_key;
self.refresh_tool_scope(abort_signal).await?;
Ok(()) Ok(())
} }
pub async fn attach_rag(&mut self, name: &str) -> Result<()> { pub async fn attach_rag(&mut self, name: &str, abort_signal: AbortSignal) -> Result<()> {
let rag_path = self.rag_file(name); let rag_path = self.rag_file(name);
if rag_path.exists() { if rag_path.exists() {
bail!( bail!(
@@ -4192,6 +4204,7 @@ impl RequestContext {
self.rag_cache().insert(key.clone(), &rag); self.rag_cache().insert(key.clone(), &rag);
self.rag = Some(rag); self.rag = Some(rag);
self.rag_key = Some(key); self.rag_key = Some(key);
self.refresh_tool_scope(abort_signal).await?;
Ok(()) Ok(())
} }
+21
View File
@@ -1,4 +1,5 @@
pub(crate) mod memory; pub(crate) mod memory;
pub(crate) mod rag_query;
pub(crate) mod skill; pub(crate) mod skill;
pub(crate) mod supervisor; pub(crate) mod supervisor;
pub(crate) mod todo; pub(crate) mod todo;
@@ -23,6 +24,7 @@ use futures_util::future;
use indexmap::IndexMap; use indexmap::IndexMap;
use indoc::formatdoc; use indoc::formatdoc;
use memory::MEMORY_FUNCTION_PREFIX; use memory::MEMORY_FUNCTION_PREFIX;
use rag_query::RAG_FUNCTION_PREFIX;
use rust_embed::Embed; use rust_embed::Embed;
use serde::{Deserialize, Serialize}; use serde::{Deserialize, Serialize};
use serde_json::{Value, json}; use serde_json::{Value, json};
@@ -495,6 +497,16 @@ impl Functions {
.extend(user_interaction::user_interaction_function_declarations()); .extend(user_interaction::user_interaction_function_declarations());
} }
pub fn append_rag_query_functions(&mut self) {
self.declarations
.extend(rag_query::rag_query_function_declarations());
}
pub fn remove_rag_query_functions(&mut self) {
self.declarations
.retain(|f| !f.name.starts_with(RAG_FUNCTION_PREFIX));
}
pub fn append_mcp_meta_functions(&mut self, mcp_servers: Vec<String>) { pub fn append_mcp_meta_functions(&mut self, mcp_servers: Vec<String>) {
let mut invoke_function_properties = IndexMap::new(); let mut invoke_function_properties = IndexMap::new();
invoke_function_properties.insert( invoke_function_properties.insert(
@@ -1252,6 +1264,15 @@ impl ToolCall {
json!({"tool_call_error": error_msg}) json!({"tool_call_error": error_msg})
}) })
} }
_ if cmd_name.starts_with(RAG_FUNCTION_PREFIX) => {
rag_query::handle_rag_tool(ctx, &cmd_name, &json_data)
.await
.unwrap_or_else(|e| {
let error_msg = format!("RAG query failed: {e}");
eprintln!("{}", muted_warning_text(&format!("⚠️ {error_msg} ⚠️")));
json!({"tool_call_error": error_msg})
})
}
_ => match run_llm_function(cmd_name, cmd_args, envs, agent_name) { _ => match run_llm_function(cmd_name, cmd_args, envs, agent_name) {
Ok(Some(contents)) => serde_json::from_str(&contents) Ok(Some(contents)) => serde_json::from_str(&contents)
.ok() .ok()
+101
View File
@@ -0,0 +1,101 @@
use super::{FunctionDeclaration, JsonSchema};
use crate::config::RequestContext;
use anyhow::{Result, anyhow};
use indexmap::IndexMap;
use serde_json::{Value, json};
pub const RAG_FUNCTION_PREFIX: &str = "rag__";
pub fn rag_query_function_declarations() -> Vec<FunctionDeclaration> {
vec![FunctionDeclaration {
name: format!("{RAG_FUNCTION_PREFIX}query"),
description: "Search the RAG knowledge base attached to this session and return \
the most relevant text chunks with their source paths. The relevant \
context has already been injected into the prompt up-front; use this \
tool to pull additional context on-demand when the initial retrieval \
does not fully answer the question. Prefer specific, keyword-rich queries."
.to_string(),
parameters: JsonSchema {
type_value: Some("object".to_string()),
properties: Some(IndexMap::from([
(
"query".to_string(),
JsonSchema {
type_value: Some("string".to_string()),
description: Some(
"Natural language search query used to retrieve relevant chunks."
.into(),
),
..Default::default()
},
),
(
"top_k".to_string(),
JsonSchema {
type_value: Some("integer".to_string()),
description: Some(
"Maximum number of chunks to return. Defaults to the RAG's \
configured top_k when omitted."
.into(),
),
..Default::default()
},
),
])),
required: Some(vec!["query".to_string()]),
..Default::default()
},
agent: false,
}]
}
pub async fn handle_rag_tool(
ctx: &mut RequestContext,
cmd_name: &str,
args: &Value,
) -> Result<Value> {
let action = cmd_name
.strip_prefix(RAG_FUNCTION_PREFIX)
.unwrap_or(cmd_name);
match action {
"query" => handle_query(ctx, args).await,
_ => Err(anyhow!("Unknown RAG action: {action}")),
}
}
async fn handle_query(ctx: &RequestContext, args: &Value) -> Result<Value> {
let rag = ctx
.rag
.clone()
.ok_or_else(|| anyhow!("No RAG is attached to this session"))?;
let query = args
.get("query")
.and_then(Value::as_str)
.ok_or_else(|| anyhow!("'query' is required"))?;
let top_k = args
.get("top_k")
.and_then(Value::as_u64)
.map(|v| v as usize)
.unwrap_or_else(|| rag.configured_top_k());
let rerank_model = rag.configured_reranker().map(|s| s.to_string());
let chunks = rag
.search_chunks(query, top_k, rerank_model.as_deref())
.await?;
let chunks_json: Vec<Value> = chunks
.into_iter()
.map(|(text, source)| json!({ "text": text, "source": source }))
.collect();
Ok(json!({
"rag_name": rag.name(),
"count": chunks_json.len(),
"chunks": chunks_json,
}))
}
+12
View File
@@ -196,6 +196,18 @@ async fn main() -> Result<()> {
return Ok(()); return Ok(());
} }
let mcp_action =
cli.mcp_list || cli.mcp_get.is_some() || cli.mcp_remove.is_some() || cli.mcp_add.is_some();
if mcp_action {
let cfg = Config::load_with_interpolation(true).await?;
let app_config = AppConfig::from_config(cfg)?;
let vault = Vault::init(&app_config)?;
mcp::manage::handle(&cli, &vault)?;
return Ok(());
}
if vault_flags { if vault_flags {
let cfg = Config::load_with_interpolation(true).await?; let cfg = Config::load_with_interpolation(true).await?;
let app_config = AppConfig::from_config(cfg)?; let app_config = AppConfig::from_config(cfg)?;
+443
View File
@@ -0,0 +1,443 @@
use crate::cli::{Cli, McpScopeArg, McpTransportArg};
use crate::config::{ensure_parent_exists, paths};
use crate::mcp::{JsonField, McpOAuthConfig, McpServer, McpServersConfig, McpTransportType};
use crate::vault::{SECRET_RE, Vault};
use anyhow::{Context, Result, anyhow, bail};
use indexmap::{IndexMap, IndexSet};
use inquire::Confirm;
use std::collections::HashSet;
use std::fs;
use std::path::{Path, PathBuf};
impl From<McpTransportArg> for McpTransportType {
fn from(value: McpTransportArg) -> Self {
match value {
McpTransportArg::Stdio => McpTransportType::Stdio,
McpTransportArg::Http => McpTransportType::Http,
McpTransportArg::Sse => McpTransportType::Sse,
}
}
}
pub fn handle(cli: &Cli, vault: &Vault) -> Result<()> {
if cli.mcp_list {
return handle_list(cli.scope);
}
if let Some(name) = &cli.mcp_get {
return handle_get(name, cli.scope);
}
if let Some(name) = &cli.mcp_remove {
return handle_remove(name, cli.scope, cli.mcp_force);
}
if let Some(name) = &cli.mcp_add {
return handle_add(cli, name, vault);
}
Ok(())
}
fn handle_list(scope: Option<McpScopeArg>) -> Result<()> {
let show_user = scope != Some(McpScopeArg::Workspace);
let show_workspace = scope != Some(McpScopeArg::User);
if show_user {
let user_path = paths::mcp_config_file();
let user_cfg = load_config_raw(&user_path)?;
println!("User ({})", user_path.display());
print_server_list(&user_cfg);
}
if show_workspace {
match paths::workspace_mcp_config_file() {
Some(ws_path) => {
let ws_cfg = load_config_raw(&ws_path)?;
if show_user {
println!();
}
println!("Workspace ({})", ws_path.display());
print_server_list(&ws_cfg);
}
None if scope == Some(McpScopeArg::Workspace) => {
println!("Workspace: no mcp.json found in current directory");
}
None => {}
}
}
Ok(())
}
fn print_server_list(cfg: &McpServersConfig) {
if cfg.mcp_servers.is_empty() {
println!(" (none)");
return;
}
let name_width = cfg.mcp_servers.keys().map(String::len).max().unwrap_or(0);
for (name, spec) in &cfg.mcp_servers {
let transport = match spec.transport_type {
McpTransportType::Stdio => "stdio",
McpTransportType::Http => "http",
McpTransportType::Sse => "sse",
};
let target = spec.url.clone().unwrap_or_else(|| {
let cmd = spec.command.clone().unwrap_or_default();
let args = spec.args.as_ref().map(|a| a.join(" ")).unwrap_or_default();
if args.is_empty() {
cmd
} else {
format!("{cmd} {args}")
}
});
println!(
" {name:<name_width$} {transport:<5} {target}",
name_width = name_width
);
}
}
fn handle_get(name: &str, scope: Option<McpScopeArg>) -> Result<()> {
let (path, cfg) = load_for_scope_or_search(name, scope)?;
let spec = cfg
.mcp_servers
.get(name)
.ok_or_else(|| anyhow!("MCP server '{name}' not found"))?;
let pretty =
serde_json::to_string_pretty(spec).context("failed to serialize MCP server config")?;
println!("# {}", path.display());
println!("{pretty}");
Ok(())
}
fn handle_remove(name: &str, scope: Option<McpScopeArg>, force: bool) -> Result<()> {
let (path, mut cfg) = load_for_scope_or_search(name, scope)?;
if !force {
let ok = Confirm::new(&format!(
"Remove MCP server '{name}' from {}?",
path.display()
))
.with_default(false)
.prompt()?;
if !ok {
println!("Aborted.");
return Ok(());
}
}
cfg.mcp_servers.shift_remove(name);
save_config(&path, &cfg)?;
println!("✓ Removed MCP server '{name}' from {}", path.display());
Ok(())
}
fn handle_add(cli: &Cli, name: &str, vault: &Vault) -> Result<()> {
validate_name(name)?;
let server = build_server(cli)?;
server.validate(name)?;
let scope = cli.scope.unwrap_or_default();
let path = write_path_for_scope(scope);
let mut cfg = load_config_raw(&path)?;
if cfg.mcp_servers.contains_key(name) && !cli.mcp_force {
let ok = Confirm::new(&format!(
"MCP server '{name}' already exists in {}. Overwrite?",
path.display()
))
.with_default(false)
.prompt()?;
if !ok {
println!("Aborted. Use --mcp-force to overwrite without prompting.");
return Ok(());
}
}
provision_secrets(cli, vault)?;
cfg.mcp_servers.insert(name.to_string(), server);
save_config(&path, &cfg)?;
println!("✓ Added MCP server '{name}' to {}", path.display());
Ok(())
}
fn validate_name(name: &str) -> Result<()> {
if name.is_empty() {
bail!("MCP server name cannot be empty");
}
if !name
.chars()
.all(|c| c.is_ascii_alphanumeric() || c == '-' || c == '_')
{
bail!("Invalid MCP server name '{name}': only letters, digits, '-', and '_' are allowed");
}
Ok(())
}
fn build_server(cli: &Cli) -> Result<McpServer> {
let has_command = !cli.mcp_command.is_empty();
let has_url = cli.url.is_some();
let transport = cli
.transport
.map(McpTransportType::from)
.unwrap_or_else(|| {
if has_command {
McpTransportType::Stdio
} else {
McpTransportType::Http
}
});
match transport {
McpTransportType::Stdio => build_stdio(cli, has_url),
McpTransportType::Http | McpTransportType::Sse => build_remote(cli, transport, has_command),
}
}
fn build_stdio(cli: &Cli, has_url: bool) -> Result<McpServer> {
if cli.mcp_command.is_empty() {
bail!(
"stdio MCP server requires a command. Pass it after `--`, e.g. \
`--mcp-add NAME -- npx some-server --flag`"
);
}
if has_url {
bail!("stdio MCP server does not accept --url");
}
if !cli.header.is_empty() {
bail!("stdio MCP server does not accept --header");
}
if cli.client_id.is_some()
|| cli.client_secret.is_some()
|| cli.callback_port.is_some()
|| cli.redirect_host.is_some()
{
bail!("stdio MCP server does not accept OAuth flags");
}
let (cmd, args) = cli.mcp_command.split_first().unwrap();
let mut env: IndexMap<String, JsonField> = IndexMap::new();
for kv in &cli.env {
let (k, v) = kv
.split_once('=')
.ok_or_else(|| anyhow!("invalid --env value '{kv}': expected KEY=VALUE"))?;
if k.is_empty() {
bail!("invalid --env value '{kv}': KEY cannot be empty");
}
env.insert(k.to_string(), JsonField::Str(v.to_string()));
}
Ok(McpServer {
transport_type: McpTransportType::Stdio,
command: Some(cmd.clone()),
args: (!args.is_empty()).then(|| args.to_vec()),
env: (!env.is_empty()).then_some(env),
cwd: cli.cwd.clone(),
url: None,
headers: None,
oauth: None,
})
}
fn build_remote(cli: &Cli, transport: McpTransportType, has_command: bool) -> Result<McpServer> {
if has_command {
bail!(
"http/sse MCP server does not accept a trailing `-- <cmd>`. Use `--url` \
to specify the endpoint."
);
}
let url = cli
.url
.clone()
.ok_or_else(|| anyhow!("http/sse MCP server requires --url <URL>"))?;
if !cli.env.is_empty() {
bail!("http/sse MCP server does not accept --env; use --header instead");
}
if cli.cwd.is_some() {
bail!("http/sse MCP server does not accept --cwd");
}
let mut headers: IndexMap<String, String> = IndexMap::new();
for h in &cli.header {
let (name, value) = h
.split_once(':')
.ok_or_else(|| anyhow!("invalid --header value '{h}': expected 'Name: Value'"))?;
let name = name.trim();
let value = value.trim_start_matches(' ');
if name.is_empty() {
bail!("invalid --header value '{h}': header name cannot be empty");
}
headers.insert(name.to_string(), value.to_string());
}
let oauth = if cli.client_id.is_some()
|| cli.client_secret.is_some()
|| cli.callback_port.is_some()
|| cli.redirect_host.is_some()
{
Some(McpOAuthConfig {
client_id: cli.client_id.clone(),
client_secret: cli.client_secret.clone(),
callback_port: cli.callback_port,
redirect_host: cli.redirect_host.clone(),
})
} else {
None
};
Ok(McpServer {
transport_type: transport,
command: None,
args: None,
env: None,
cwd: None,
url: Some(url),
headers: (!headers.is_empty()).then_some(headers),
oauth,
})
}
fn provision_secrets(cli: &Cli, vault: &Vault) -> Result<()> {
let mut sources: Vec<&str> = Vec::new();
if let Some(s) = cli.url.as_deref() {
sources.push(s);
}
if let Some(s) = cli.client_secret.as_deref() {
sources.push(s);
}
if let Some(s) = cli.client_id.as_deref() {
sources.push(s);
}
if let Some(s) = cli.redirect_host.as_deref() {
sources.push(s);
}
if let Some(s) = cli.cwd.as_deref() {
sources.push(s);
}
sources.extend(cli.env.iter().map(String::as_str));
sources.extend(cli.header.iter().map(String::as_str));
let mut needed: IndexSet<String> = IndexSet::new();
for value in sources {
for caps in SECRET_RE.captures_iter(value).filter_map(Result::ok) {
if let Some(m) = caps.get(1) {
needed.insert(m.as_str().trim().to_string());
}
}
}
if needed.is_empty() {
return Ok(());
}
let existing: HashSet<String> = vault.list_secrets(false)?.into_iter().collect();
for name in needed {
if existing.contains(&name) {
continue;
}
eprintln!("Value references vault secret {{{{ {name} }}}} which is not stored yet.");
let ok = Confirm::new(&format!("Add '{name}' to the vault now?"))
.with_default(true)
.prompt()?;
if !ok {
bail!(
"Vault secret '{name}' is required by the config; aborting. \
Add it later with `coyote --add-secret {name}`."
);
}
vault.add_secret(&name)?;
}
Ok(())
}
fn load_for_scope_or_search(
name: &str,
scope: Option<McpScopeArg>,
) -> Result<(PathBuf, McpServersConfig)> {
if let Some(s) = scope {
let path = match s {
McpScopeArg::User => paths::mcp_config_file(),
McpScopeArg::Workspace => paths::workspace_mcp_config_file()
.ok_or_else(|| anyhow!("no workspace mcp.json found in the current directory"))?,
};
let cfg = load_config_raw(&path)?;
if !cfg.mcp_servers.contains_key(name) {
bail!(
"MCP server '{name}' not found in {} scope ({})",
scope_label(s),
path.display()
);
}
return Ok((path, cfg));
}
let user_path = paths::mcp_config_file();
let user_cfg = load_config_raw(&user_path)?;
if user_cfg.mcp_servers.contains_key(name) {
return Ok((user_path, user_cfg));
}
if let Some(ws_path) = paths::workspace_mcp_config_file() {
let ws_cfg = load_config_raw(&ws_path)?;
if ws_cfg.mcp_servers.contains_key(name) {
return Ok((ws_path, ws_cfg));
}
}
bail!("MCP server '{name}' not found in any scope");
}
fn write_path_for_scope(scope: McpScopeArg) -> PathBuf {
match scope {
McpScopeArg::User => paths::mcp_config_file(),
McpScopeArg::Workspace => paths::workspace_mcp_config_file()
.unwrap_or_else(|| paths::workspace_config_dir().join("mcp.json")),
}
}
fn scope_label(scope: McpScopeArg) -> &'static str {
match scope {
McpScopeArg::User => "user",
McpScopeArg::Workspace => "workspace",
}
}
fn load_config_raw(path: &Path) -> Result<McpServersConfig> {
if !path.exists() {
return Ok(McpServersConfig {
mcp_servers: IndexMap::new(),
});
}
let raw = fs::read_to_string(path)
.with_context(|| format!("failed to read MCP config at {}", path.display()))?;
if raw.trim().is_empty() {
return Ok(McpServersConfig {
mcp_servers: IndexMap::new(),
});
}
serde_json::from_str(&raw)
.with_context(|| format!("failed to parse MCP config at {}", path.display()))
}
fn save_config(path: &Path, config: &McpServersConfig) -> Result<()> {
ensure_parent_exists(path)?;
let serialized =
serde_json::to_string_pretty(config).context("failed to serialize MCP config")?;
let tmp = path.with_extension("json.tmp");
fs::write(&tmp, &serialized)
.with_context(|| format!("failed to write temporary MCP config at {}", tmp.display()))?;
fs::rename(&tmp, path)
.with_context(|| format!("failed to finalize MCP config at {}", path.display()))?;
Ok(())
}
+3
View File
@@ -1,3 +1,4 @@
pub(crate) mod manage;
pub(crate) mod oauth; pub(crate) mod oauth;
mod sse_transport; mod sse_transport;
@@ -62,6 +63,8 @@ pub(crate) struct McpServersConfig {
pub(crate) struct McpOAuthConfig { pub(crate) struct McpOAuthConfig {
#[serde(rename = "clientId", skip_serializing_if = "Option::is_none")] #[serde(rename = "clientId", skip_serializing_if = "Option::is_none")]
pub client_id: Option<String>, pub client_id: Option<String>,
#[serde(rename = "clientSecret", skip_serializing_if = "Option::is_none")]
pub client_secret: Option<String>,
#[serde(rename = "callbackPort", skip_serializing_if = "Option::is_none")] #[serde(rename = "callbackPort", skip_serializing_if = "Option::is_none")]
pub callback_port: Option<u16>, pub callback_port: Option<u16>,
#[serde(rename = "redirectHost", skip_serializing_if = "Option::is_none")] #[serde(rename = "redirectHost", skip_serializing_if = "Option::is_none")]
+14
View File
@@ -856,6 +856,20 @@ impl Rag {
Ok((embeddings, sources, ids)) Ok((embeddings, sources, ids))
} }
pub async fn search_chunks(
&self,
text: &str,
top_k: usize,
rerank_model: Option<&str>,
) -> Result<Vec<(String, String)>> {
let results = self.hybrid_search(text, top_k, rerank_model).await?;
Ok(results
.into_iter()
.map(|(id, content)| (content, self.resolve_source(&id)))
.collect())
}
pub async fn search_with_template( pub async fn search_with_template(
&self, &self,
app: &AppConfig, app: &AppConfig,
+1 -1
View File
@@ -892,7 +892,7 @@ pub async fn run_repl_command(
".rag" => match split_first_arg(args) { ".rag" => match split_first_arg(args) {
Some(("attach", rest)) => match rest { Some(("attach", rest)) => match rest {
Some(name) if !name.trim().is_empty() => { Some(name) if !name.trim().is_empty() => {
ctx.attach_rag(name.trim()).await?; ctx.attach_rag(name.trim(), abort_signal.clone()).await?;
} }
_ => println!("Usage: .rag attach <name>"), _ => println!("Usage: .rag attach <name>"),
}, },
+318 -7
View File
@@ -1,9 +1,10 @@
use std::env; use std::env;
use std::fs; use std::fs;
use std::fs::{read_dir, read_to_string}; use std::fs::{read_dir, read_to_string};
use std::io;
use std::path::{Path, PathBuf}; use std::path::{Path, PathBuf};
use anyhow::{Context, Result}; use anyhow::{Context, Result, anyhow, bail};
use serde_yaml::Value; use serde_yaml::Value;
use sha2::{Digest, Sha256}; use sha2::{Digest, Sha256};
@@ -12,6 +13,7 @@ use crate::config::paths;
const SBX_MIXIN_FILE_NAME: &str = "sbx-mixin.yaml"; const SBX_MIXIN_FILE_NAME: &str = "sbx-mixin.yaml";
const SBX_MIXIN_FILE_SUFFIX: &str = ".sbx-mixin.yaml"; const SBX_MIXIN_FILE_SUFFIX: &str = ".sbx-mixin.yaml";
const KIT_SPEC_FILE_NAME: &str = "spec.yaml"; const KIT_SPEC_FILE_NAME: &str = "spec.yaml";
const MIXIN_FILES_DIR_NAME: &str = "files";
#[derive(Debug, Clone)] #[derive(Debug, Clone)]
pub struct DiscoveredMixin { pub struct DiscoveredMixin {
@@ -34,33 +36,152 @@ impl DiscoveredMixin {
pub fn wrap_mixin_as_kit(mixin_path: &Path) -> Result<PathBuf> { pub fn wrap_mixin_as_kit(mixin_path: &Path) -> Result<PathBuf> {
let bytes = fs::read(mixin_path) let bytes = fs::read(mixin_path)
.with_context(|| format!("Failed to read sbx mixin {}", mixin_path.display()))?; .with_context(|| format!("Failed to read sbx mixin {}", mixin_path.display()))?;
wrap_mixin_bytes_as_kit(&bytes, &mixin_path.display().to_string()) let label = mixin_path.display().to_string();
let files = mixin_path
.parent()
.map(|p| p.join(MIXIN_FILES_DIR_NAME))
.filter(|p| p.is_dir())
.map(|dir| collect_staged_files(&dir))
.transpose()?
.unwrap_or_default();
stage_kit(&bytes, &files, &label)
} }
pub fn wrap_mixin_bytes_as_kit(bytes: &[u8], label: &str) -> Result<PathBuf> { pub fn wrap_mixin_bytes_as_kit(bytes: &[u8], label: &str) -> Result<PathBuf> {
stage_kit(bytes, &[], label)
}
struct StagedFile {
relpath: PathBuf,
mode: u32,
bytes: Vec<u8>,
}
fn stage_kit(spec_bytes: &[u8], files: &[StagedFile], label: &str) -> Result<PathBuf> {
let mut hasher = Sha256::new(); let mut hasher = Sha256::new();
hasher.update(bytes); hasher.update(spec_bytes);
for f in files {
let rel_str = f.relpath.to_str().ok_or_else(|| {
anyhow!(
"Non-UTF-8 path inside mixin {MIXIN_FILES_DIR_NAME}/: {}",
f.relpath.display()
)
})?;
hasher.update(b"\0COYOTE_MIXIN_FILE\0");
hasher.update((rel_str.len() as u64).to_le_bytes());
hasher.update(rel_str.as_bytes());
hasher.update(f.mode.to_le_bytes());
hasher.update((f.bytes.len() as u64).to_le_bytes());
hasher.update(&f.bytes);
}
let hash = format!("{:x}", hasher.finalize()); let hash = format!("{:x}", hasher.finalize());
let kit_dir = paths::sbx_mixin_kits_dir().join(&hash); let kit_dir = paths::sbx_mixin_kits_dir().join(&hash);
let spec_path = kit_dir.join(KIT_SPEC_FILE_NAME); let spec_path = kit_dir.join(KIT_SPEC_FILE_NAME);
let files_dst = kit_dir.join(MIXIN_FILES_DIR_NAME);
if let Ok(existing) = fs::read(&spec_path) let spec_matches = fs::read(&spec_path).is_ok_and(|existing| existing == spec_bytes);
&& existing == bytes let files_ready = files.is_empty() || files_dst.is_dir();
{ if spec_matches && files_ready {
return Ok(kit_dir); return Ok(kit_dir);
} }
fs::create_dir_all(&kit_dir) fs::create_dir_all(&kit_dir)
.with_context(|| format!("Failed to create mixin kit dir {}", kit_dir.display()))?; .with_context(|| format!("Failed to create mixin kit dir {}", kit_dir.display()))?;
fs::write(&spec_path, bytes) fs::write(&spec_path, spec_bytes)
.with_context(|| format!("Failed to write {}", spec_path.display()))?; .with_context(|| format!("Failed to write {}", spec_path.display()))?;
if !files.is_empty() {
if files_dst.exists() {
fs::remove_dir_all(&files_dst).with_context(|| {
format!(
"Failed to clear stale mixin files at {}",
files_dst.display()
)
})?;
}
for f in files {
let dst = files_dst.join(&f.relpath);
if let Some(parent) = dst.parent() {
fs::create_dir_all(parent)
.with_context(|| format!("Failed to create dir {}", parent.display()))?;
}
fs::write(&dst, &f.bytes)
.with_context(|| format!("Failed to write staged mixin file {}", dst.display()))?;
#[cfg(unix)]
{
use std::os::unix::fs::PermissionsExt;
fs::set_permissions(&dst, fs::Permissions::from_mode(f.mode))
.with_context(|| format!("Failed to set mode on {}", dst.display()))?;
}
}
}
debug!("Wrapped mixin {label} as kit at {}", kit_dir.display()); debug!("Wrapped mixin {label} as kit at {}", kit_dir.display());
Ok(kit_dir) Ok(kit_dir)
} }
fn collect_staged_files(root: &Path) -> Result<Vec<StagedFile>> {
let mut out = Vec::new();
walk_staged_files(root, Path::new(""), &mut out)?;
Ok(out)
}
fn walk_staged_files(abs_dir: &Path, rel_dir: &Path, out: &mut Vec<StagedFile>) -> Result<()> {
let rd = fs::read_dir(abs_dir)
.with_context(|| format!("Failed to read mixin files dir {}", abs_dir.display()))?;
let mut entries: Vec<_> = rd
.collect::<io::Result<Vec<_>>>()
.with_context(|| format!("Failed to iterate mixin files dir {}", abs_dir.display()))?;
entries.sort_by_key(|e| e.file_name());
for entry in entries {
let file_type = entry
.file_type()
.with_context(|| format!("Failed to stat {}", entry.path().display()))?;
let abs = entry.path();
let rel = rel_dir.join(entry.file_name());
if file_type.is_symlink() {
bail!(
"Symlinks are not allowed inside a mixin {MIXIN_FILES_DIR_NAME}/ tree: {}",
abs.display()
);
}
if file_type.is_dir() {
walk_staged_files(&abs, &rel, out)?;
} else if file_type.is_file() {
let bytes = fs::read(&abs)
.with_context(|| format!("Failed to read staged mixin file {}", abs.display()))?;
let mode = staged_file_mode(&entry)?;
out.push(StagedFile {
relpath: rel,
mode,
bytes,
});
}
}
Ok(())
}
#[cfg(unix)]
fn staged_file_mode(entry: &fs::DirEntry) -> Result<u32> {
use std::os::unix::fs::PermissionsExt;
let meta = entry
.metadata()
.with_context(|| format!("Failed to stat {}", entry.path().display()))?;
Ok(meta.permissions().mode() & 0o777)
}
#[cfg(not(unix))]
fn staged_file_mode(_entry: &fs::DirEntry) -> Result<u32> {
Ok(0o644)
}
pub fn discover() -> Result<Vec<DiscoveredMixin>> { pub fn discover() -> Result<Vec<DiscoveredMixin>> {
let mut out = Vec::new(); let mut out = Vec::new();
@@ -556,6 +677,196 @@ network:
"kit_path should not return the original file path" "kit_path should not return the original file path"
); );
} }
fn write_staged_file(mixin: &Path, rel: &str, content: &[u8]) {
let dst = mixin.parent().unwrap().join(MIXIN_FILES_DIR_NAME).join(rel);
fs::create_dir_all(dst.parent().unwrap()).unwrap();
fs::write(&dst, content).unwrap();
}
#[test]
#[serial]
fn wrap_mixin_as_kit_copies_sibling_files_tree_into_kit() {
let _guard = TestCacheDirGuard::new();
let mixin = write_mixin("files-copy", "kind: mixin\nname: probe\n");
write_staged_file(&mixin, "home/hello.md", b"# hello\n");
write_staged_file(&mixin, "home/nested/deep.txt", b"deep\n");
let kit_dir = wrap_mixin_as_kit(&mixin).unwrap();
assert!(kit_dir.join("spec.yaml").exists());
let files_root = kit_dir.join(MIXIN_FILES_DIR_NAME);
assert!(files_root.is_dir(), "kit dir must contain a files/ tree");
assert_eq!(
fs::read(files_root.join("home/hello.md")).unwrap(),
b"# hello\n"
);
assert_eq!(
fs::read(files_root.join("home/nested/deep.txt")).unwrap(),
b"deep\n"
);
}
#[test]
#[serial]
fn wrap_mixin_as_kit_hash_changes_when_a_staged_file_is_edited() {
let _guard = TestCacheDirGuard::new();
let mixin = write_mixin("files-hash-content", "kind: mixin\nname: probe\n");
write_staged_file(&mixin, "home/note.md", b"before\n");
let kit_before = wrap_mixin_as_kit(&mixin).unwrap();
write_staged_file(&mixin, "home/note.md", b"after\n");
let kit_after = wrap_mixin_as_kit(&mixin).unwrap();
assert_ne!(
kit_before, kit_after,
"editing a staged file must invalidate the kit hash"
);
assert_eq!(
fs::read(kit_after.join("files/home/note.md")).unwrap(),
b"after\n"
);
}
#[test]
#[serial]
fn wrap_mixin_as_kit_hash_changes_when_a_staged_file_is_added() {
let _guard = TestCacheDirGuard::new();
let mixin = write_mixin("files-hash-added", "kind: mixin\nname: probe\n");
write_staged_file(&mixin, "home/one.md", b"one\n");
let kit_before = wrap_mixin_as_kit(&mixin).unwrap();
write_staged_file(&mixin, "home/two.md", b"two\n");
let kit_after = wrap_mixin_as_kit(&mixin).unwrap();
assert_ne!(
kit_before, kit_after,
"adding a staged file must invalidate the kit hash"
);
}
#[test]
#[serial]
fn wrap_mixin_as_kit_hash_unchanged_when_no_files_dir() {
let _guard = TestCacheDirGuard::new();
let content = "kind: mixin\nname: legacy\n";
let mixin = write_mixin("legacy-no-files", content);
let with_helper = wrap_mixin_as_kit(&mixin).unwrap();
let bytes_only = wrap_mixin_bytes_as_kit(content.as_bytes(), "legacy").unwrap();
assert_eq!(
with_helper, bytes_only,
"mixins without a sibling files/ must keep the legacy bytes-only hash to reuse existing cache dirs"
);
}
#[test]
#[serial]
fn wrap_mixin_as_kit_ignores_sibling_files_that_is_not_a_directory() {
let _guard = TestCacheDirGuard::new();
let content = "kind: mixin\nname: probe\n";
let mixin = write_mixin("files-not-a-dir", content);
fs::write(mixin.parent().unwrap().join(MIXIN_FILES_DIR_NAME), b"decoy").unwrap();
let wrapped = wrap_mixin_as_kit(&mixin).unwrap();
let bytes_only = wrap_mixin_bytes_as_kit(content.as_bytes(), "probe").unwrap();
assert_eq!(
wrapped, bytes_only,
"a regular file named files must be ignored, not staged"
);
assert!(!wrapped.join(MIXIN_FILES_DIR_NAME).exists());
}
#[test]
#[serial]
fn wrap_mixin_as_kit_rebuilds_files_when_cache_dir_missing_files_tree() {
let _guard = TestCacheDirGuard::new();
let mixin = write_mixin("files-rebuild", "kind: mixin\nname: probe\n");
write_staged_file(&mixin, "home/hello.md", b"hi\n");
let kit_dir = wrap_mixin_as_kit(&mixin).unwrap();
let files_dst = kit_dir.join(MIXIN_FILES_DIR_NAME);
fs::remove_dir_all(&files_dst).unwrap();
assert!(!files_dst.exists());
let kit_again = wrap_mixin_as_kit(&mixin).unwrap();
assert_eq!(kit_again, kit_dir, "kit path is content-addressed");
assert!(
files_dst.is_dir(),
"a partial cache (spec present, files/ missing) must be rebuilt"
);
assert_eq!(fs::read(files_dst.join("home/hello.md")).unwrap(), b"hi\n");
}
#[test]
#[serial]
fn wrap_mixin_as_kit_deterministic_with_staged_files() {
let _guard = TestCacheDirGuard::new();
let content = "kind: mixin\nname: probe\n";
let mixin_one = write_mixin("determ-1", content);
write_staged_file(&mixin_one, "home/note.md", b"same\n");
let mixin_two = write_mixin("determ-2", content);
write_staged_file(&mixin_two, "home/note.md", b"same\n");
let kit_a = wrap_mixin_as_kit(&mixin_one).unwrap();
let kit_b = wrap_mixin_as_kit(&mixin_two).unwrap();
assert_eq!(
kit_a, kit_b,
"identical spec+files must produce the same content-addressed kit dir"
);
}
#[cfg(unix)]
#[test]
#[serial]
fn wrap_mixin_as_kit_rejects_symlinks_inside_files_tree() {
use std::os::unix::fs::symlink;
let _guard = TestCacheDirGuard::new();
let mixin = write_mixin("files-symlink", "kind: mixin\nname: probe\n");
let files_dir = mixin.parent().unwrap().join(MIXIN_FILES_DIR_NAME);
fs::create_dir_all(&files_dir).unwrap();
let target = files_dir.join("target.txt");
fs::write(&target, b"real").unwrap();
symlink(&target, files_dir.join("link.txt")).unwrap();
let err = wrap_mixin_as_kit(&mixin).unwrap_err();
let msg = format!("{err:#}");
assert!(
msg.contains("Symlinks are not allowed"),
"expected symlink rejection, got: {msg}"
);
}
#[cfg(unix)]
#[test]
#[serial]
fn wrap_mixin_as_kit_preserves_executable_bit() {
use std::os::unix::fs::PermissionsExt;
let _guard = TestCacheDirGuard::new();
let mixin = write_mixin("files-exec", "kind: mixin\nname: probe\n");
write_staged_file(&mixin, "bin/run.sh", b"#!/bin/sh\necho hi\n");
let src = mixin
.parent()
.unwrap()
.join(MIXIN_FILES_DIR_NAME)
.join("bin/run.sh");
fs::set_permissions(&src, fs::Permissions::from_mode(0o755)).unwrap();
let kit_dir = wrap_mixin_as_kit(&mixin).unwrap();
let dst = kit_dir.join("files/bin/run.sh");
let mode = fs::metadata(&dst).unwrap().permissions().mode() & 0o777;
assert_eq!(
mode, 0o755,
"executable bit must survive the copy into the kit dir"
);
}
} }
#[test] #[test]