| name | test-hooks |
| description | Hook health check and test runner — verify hooks work, hook performance, diagnose hook failures. Trigger: hook health check, verify hooks, test hooks, hook performance. |
/test-hooks — Hook Health Check and Test Runner
Run all registered hooks with synthetic payloads to verify they work correctly.
Arguments
/test-hooks — Run all hooks/test-hooks --event <EventName> — Test only hooks for a specific event (e.g., Stop, SessionStart)/test-hooks --hook <filename> — Test a specific hook by filename/test-hooks --verbose — Show stdout/stderr from each hook
Workflow
- Run the test runner:
node ~/.claude/scripts/test-hooks.js [args]
-
Review the output. For each hook, the runner reports:
[OK] PASS with timing in ms[!!] FAIL with error message
-
For any failures:
- Read the failing hook file
- Check if the command path exists:
ls -la <path>
- Check if dependencies are available (e.g.,
which python3, which node)
- Common fixes: missing file, wrong path separator, missing module
-
Report summary to user:
- Total: N passed, N failed
- By event: breakdown per event type
- Slow hooks: any taking >1s
Notes
- Hooks receive synthetic payloads, not real session data
- Exit code 2 with stderr is valid (hook intentionally blocking) — counted as PASS
- Timeout: 5s per hook (hooks should complete in <200ms normally)
- K-LEAN hooks (
kln-hook-*) require K-LEAN to be installed
- The test runner is at
~/.claude/scripts/test-hooks.js
