Agent skill
handoff
Manages workflow transitions and completion reports in 2-Agent workflow. Use when user mentions '/handoff', completion report, handoff to Cursor/OpenCode, auto-fix, or reporting to PM. Do NOT load for: casual completion statements, progress chat, informal status updates.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/handoff-chachamaru127-claude-code-harness
SKILL.md
Handoff Skill
PM-実装役間のハンドオフとワークフロー遷移を管理するスキル。
機能詳細
| 機能 | 詳細 |
|---|---|
| PM→実装役 | See references/handoff-to-impl.md |
| 実装役→PM | See references/handoff-to-pm.md |
| レビュー指摘自動修正 | See references/auto-fixing.md |
| コミット実行 | See references/execute-commit.md |
Quick Reference
- "Cursor に完了報告を書いて" →
/handoff cursor - "OpenCode にハンドオフ" →
/handoff opencode - "レビュー指摘を自動修正" →
/handoff auto-fix - "変更内容とテスト結果を含めて" → Includes git diff and test results
Prerequisites
This command should only run after harness-review APPROVE
| Condition | Required | Check Method |
|---|---|---|
| harness-review completed | Yes | Review result is APPROVE |
| No Critical/High issues | Yes | All fixed |
| Implementation complete | Yes | Plans.md tasks completed |
Why handoff requires review approval:
- PM receives unreviewed changes otherwise
- Quality not assured
- Breaks
/workflow (implement → review → fix → OK → handoff)
Usage
/handoff cursor # Handoff to Cursor
/handoff opencode # Handoff to OpenCode
Execution Flow
Step 1: Identify Completed Tasks
- Check Plans.md checkboxes
- Summarize work done
Step 2: Update Plans.md
# Before
- [ ] Task name `pm:依頼中`
# After
- [x] Task name `cc:完了` (YYYY-MM-DD)
Step 3: Gather Changes
git status -sb
git diff --stat
Step 4: Check CI/CD (if applicable)
gh run list --limit 3
Step 5: Generate Report
Output Format
## Completion Report
### Summary
- (1-3 lines describing what was done)
### Completed Tasks
- **Task Name**: [Task description]
### Changed Files
| File | Changes |
|------|---------|
| `path/to/file1` | [Summary] |
| `path/to/file2` | [Summary] |
### Verification Results
- [x] Build success
- [x] Tests passed
- [x] Manual verification complete
### Risks / Notes
- (If any)
### Next Actions (for PM)
1. [ ] [What PM should do next]
2. [ ] [Optional items]
/work Integration Flow
/work execution
↓
Phase 1: Parallel implementation
↓
Phase 2: harness-review loop
├── NG (Critical/High) → Fix → Re-review
└── OK (APPROVE) → Phase 3
↓
Phase 3: Auto-commit (if configured)
↓
Phase 4: This skill runs ← First time handoff happens
/workautomatically calls this skill in Phase 4. When running manually, always do so after harness-review APPROVE.
Target-Specific Notes
Handoff to Cursor
- Plans.md markers use
cc:完了(Japanese) - Report format optimized for Cursor PM workflow
- Includes context for
/review-cc-workcommand
Handoff to OpenCode
- Similar format to Cursor
- Compatible with OpenCode's command structure
- Works with multi-LLM development workflow
Related Skills
work- Main implementation workflowharness-review- Code review2agent- 2-Agent workflow setup
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?