| name | tf-review |
| description | Review REVIEW-status tasks as PM, check code quality, then approve or reject. |
| disable-model-invocation | true |
| allowed-tools | Bash, Read, Glob, Grep |
Marblo PM Review
Review the REVIEW tasks agents submitted.
Check code quality and approve (DONE) or reject (revert to TODO).
⛔ Required rule: Marblo MCP only
Never use the Claude Code built-in tools (TaskCreate, TaskList, TaskUpdate, TaskGet).
Task lookup/status-change MUST use the Marblo MCP tools:
get_all_tasks, update_task_status, add_activity, get_task_activities
Step 1: Fetch tasks awaiting review
Filter get_all_tasks to only REVIEW status.
👀 Awaiting review: {n}
━━━━━━━━━━━━━
1. TASK-002: User API (backend)
Submitted: 30 min ago | Owner: backend-agent
2. TASK-004: URL input screen (frontend)
Submitted: 15 min ago | Owner: frontend-agent
Step 2: Review each task
For each task, check the following:
2-1. Check the activity log
Use get_task_activities to understand what the agent did.
2-2. Check the code
Read and review the files listed in the task's scope field.
2-3. Review checklist
□ Feature completeness: are all requirements from the task description implemented?
□ Code quality: did it follow the skill file's rules (framework, patterns)?
□ Tests included: is there test code and does it pass?
□ Error handling: are exceptional cases handled properly?
□ Scope compliance: did it avoid modifying files outside scope?
□ Security: no vulnerabilities like SQL injection, XSS?
2-4. Verdict
Show the review result to the user:
📝 TASK-002: User API review result
━━━━━━━━━━━━━━━━━━━━━━━━━
✅ Feature completeness: signup, login, profile lookup — OK
✅ Code quality: FastAPI + Pydantic v2 — OK
✅ Tests: 5/5 pass — OK
⚠️ Error handling: no password validation
✅ Scope compliance: OK
Verdict: approve or reject?
Step 3: Approve / reject
Approve
Reject
After review
After handling all REVIEWs, show the status:
📊 Review complete:
Approved: {n}
Rejected: {n}
Newly available tasks: {list}
Next action: /tf-work or re-run the agents