refactor: fold BreadcrumbSource into KNOWN_AGENTS (#4)

## Summary

- Moves `breadcrumb_dir` and `breadcrumb_ext` fields directly into the
`Agent` struct, eliminating the separate `BreadcrumbSource` struct and
`SOURCES` constant
- Removes the `find_agent()` helper that matched agents by email prefix
— no longer needed since we iterate `KNOWN_AGENTS` directly
- Makes it easier to add breadcrumb support for new agents by keeping
all agent metadata in one place
- Simplifies the breadcrumb fallback section in README

Addresses
#3 (comment)

## Test plan

- [x] `cargo test` — all 16 tests pass
- [x] `./scripts/aittributor --debug` — breadcrumb fallback still works
correctly

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Author: baldawarishi

Cargo.lock

@@ -60,9 +60,4 @@ ln -s /usr/local/bin/aittributor .git/hooks/prepare-commit-msg
 
 ## Breadcrumb fallback
 
-The commit hook relies on detecting a running AI agent process. If the agent exits before you commit, the hook won't find it. As a fallback, aittributor checks agent-specific state files ("breadcrumbs") left behind by agents:
-
-- **Claude Code**: checks `~/.claude/projects/` for recent session files matching the current repo
-- **Codex**: checks `~/.codex/sessions/` for recent session files whose `cwd` is within the current git root (monorepo subdirectories match; sibling repo names like `repo` and `repo2` do not)
-
-Files modified within the last 2 hours are considered recent. No additional setup is required — these directories are created automatically by the agents themselves.
+If the AI agent exits before you commit, aittributor falls back to checking agent-specific state files to detect recently active agents. This only works when state files are available. No additional setup is required. 

README.md

@@ -4,63 +4,87 @@ pub struct Agent {
     pub process_names: &'static [&'static str],
     pub env_vars: &'static [(&'static str, &'static str)],
     pub email: &'static str,
+    pub breadcrumb_dir: Option<&'static str>,
+    pub breadcrumb_ext: Option<&'static str>,
 }
 
 pub const KNOWN_AGENTS: &[Agent] = &[
     Agent {
         process_names: &["claude"],
         env_vars: &[],
         email: "Claude Code <noreply@anthropic.com>",
+        breadcrumb_dir: Some(".claude/projects"),
+        breadcrumb_ext: Some("jsonl"),
     },
     Agent {
         process_names: &["goose"],
         env_vars: &[],
         email: "Goose <opensource@block.xyz>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
     Agent {
         process_names: &["cursor", "cursor-agent"],
         env_vars: &[],
         email: "Cursor <noreply@cursor.com>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
     Agent {
         process_names: &["aider"],
         env_vars: &[],
         email: "Aider <noreply@aider.chat>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
     Agent {
         process_names: &["windsurf"],
         env_vars: &[],
         email: "Windsurf <noreply@codeium.com>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
     Agent {
         process_names: &["codex"],
         env_vars: &[],
         email: "Codex <noreply@openai.com>",
+        breadcrumb_dir: Some(".codex/sessions"),
+        breadcrumb_ext: Some("jsonl"),
     },
     Agent {
         process_names: &["copilot-agent"],
         env_vars: &[],
         email: "GitHub Copilot <noreply@github.com>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
     Agent {
         process_names: &["amazon-q", "q"],
         env_vars: &[],
         email: "Amazon Q Developer <noreply@amazon.com>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
     Agent {
         process_names: &["amp"],
         env_vars: &[],
         email: "Amp <amp@ampcode.com>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
     Agent {
         process_names: &[],
         env_vars: &[("CLINE_ACTIVE", "true")],
         email: "Cline <noreply@cline.bot>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
     Agent {
         process_names: &["gemini"],
         env_vars: &[],
         email: "Gemini CLI Agent <gemini-cli-agent@google.com>",
+        breadcrumb_dir: None,
+        breadcrumb_ext: None,
     },
 ];
 

src/agent.rs

@@ -10,28 +10,6 @@ const CUTOFF_SECS: u64 = 2 * 60 * 60; // 2 hours as a rough approximation
 /// Maximum number of lines to read from a session file when looking for "cwd".
 const MAX_LINES_TO_SCAN: usize = 5;
 
-struct BreadcrumbSource {
-    /// Prefix to match against Agent.email in KNOWN_AGENTS
-    email_prefix: &'static str,
-    /// Base directory relative to $HOME (e.g. ".claude/projects")
-    base_dir: &'static str,
-    /// File extension to look for (without dot)
-    file_ext: &'static str,
-}
-
-const SOURCES: &[BreadcrumbSource] = &[
-    BreadcrumbSource {
-        email_prefix: "Claude Code",
-        base_dir: ".claude/projects",
-        file_ext: "jsonl",
-    },
-    BreadcrumbSource {
-        email_prefix: "Codex",
-        base_dir: ".codex/sessions",
-        file_ext: "jsonl",
-    },
-];
-
 fn home_dir() -> Option<String> {
     std::env::var("HOME").ok()
 }
@@ -46,10 +24,6 @@ fn has_extension(path: &Path, ext: &str) -> bool {
     path.extension().and_then(|e| e.to_str()) == Some(ext)
 }
 
-fn find_agent(email_prefix: &str) -> Option<&'static Agent> {
-    KNOWN_AGENTS.iter().find(|a| a.email.starts_with(email_prefix))
-}
-
 fn extract_cwd_from_json(line: &str) -> Option<&str> {
     // Simple string extraction: find "cwd":"<value>"
     let marker = "\"cwd\":\"";
@@ -116,36 +90,37 @@ fn find_session_file_with_cwd(dir: &Path, ext: &str, repo_path: &Path, cutoff: S
     false
 }
 
-fn check_source(
-    source: &BreadcrumbSource,
-    repo_path: &Path,
-    cutoff: SystemTime,
-    debug: bool,
-) -> Option<&'static Agent> {
-    let home = home_dir()?;
-    let base = Path::new(&home).join(source.base_dir);
+fn check_source(agent: &'static Agent, repo_path: &Path, cutoff: SystemTime, debug: bool) -> bool {
+    let breadcrumb_dir = match agent.breadcrumb_dir {
+        Some(d) => d,
+        None => return false,
+    };
+    let breadcrumb_ext = agent.breadcrumb_ext.unwrap_or("jsonl");
+
+    let home = match home_dir() {
+        Some(h) => h,
+        None => return false,
+    };
+    let base = Path::new(&home).join(breadcrumb_dir);
 
     if debug {
-        eprintln!("  {} breadcrumb dir: {}", source.email_prefix, base.display());
+        eprintln!("  {} breadcrumb dir: {}", agent.email, base.display());
     }
 
     if !base.is_dir() {
         if debug {
             eprintln!("    Not found");
         }
-        return None;
+        return false;
     }
 
-    let matched = find_session_file_with_cwd(&base, source.file_ext, repo_path, cutoff, debug);
+    let matched = find_session_file_with_cwd(&base, breadcrumb_ext, repo_path, cutoff, debug);
 
-    if matched {
-        find_agent(source.email_prefix)
-    } else {
-        if debug {
-            eprintln!("    No match for {}", source.email_prefix);
-        }
-        None
+    if !matched && debug {
+        eprintln!("    No match for {}", agent.email);
     }
+
+    matched
 }
 
 pub fn detect_agents_from_breadcrumbs(repo_path: &Path, debug: bool) -> Vec<&'static Agent> {
@@ -156,8 +131,8 @@ pub fn detect_agents_from_breadcrumbs(repo_path: &Path, debug: bool) -> Vec<&'st
         eprintln!("\n=== Breadcrumb Fallback ===");
     }
 
-    for source in SOURCES {
-        if let Some(agent) = check_source(source, repo_path, cutoff, debug) {
+    for agent in KNOWN_AGENTS {
+        if check_source(agent, repo_path, cutoff, debug) {
             agents.push(agent);
         }
     }