loop-executor

name	loop-executor
description	Implement tasks from the plan. Evaluate-Loop Step 3.
arguments	[{"name":"track_id","description":"The track ID to execute","required":true}]
user_invocable	true

/loop-executor — Execute Plan Tasks

Evaluate-Loop Step 3: Implement the tasks defined in the plan.

Usage

/loop-executor <track-id>

Your Task

You ARE the loop-executor agent. Implement each pending task:

1. Read plan.md: Find all [ ] tasks (pending), skip [x] tasks (done)

2. For each task:

   • Read task description and acceptance criteria
   • Read existing files mentioned in the task
   • Implement the code changes following project patterns
   • Run verification (npm run build, npm run typecheck)
   • Create git commit with descriptive message
   • Immediately mark [x] in plan.md with commit SHA
   • Update metadata.json checkpoint

3. After all tasks:

   • Set current_step = "EVALUATE_EXECUTION"
   • Set step_status = "NOT_STARTED"

Commit Format

feat(track-id): Task 1.1 - Create base component

- Created src/components/foo.tsx
- Added TypeScript types

Co-Authored-By: Claude <[e-mail]>

Plan.md Update Format

- [x] Task 1.1: Create base component <!-- abc1234 -->
  - Created src/components/foo.tsx
  - Added TypeScript types

Output

TASKS COMPLETED: X/Y
COMMITS: abc1234, def5678, ...
VERDICT: PASS

Message Bus

echo "PASS" > .message-bus/events/EXECUTE_COMPLETE_{track_id}.event

Reference

Full agent instructions: .claude/agents/loop-executor.md