Agent skill
Review Coordination
Use when coordinating code reviews, agent reviews, and approval workflows across Jira and Harness
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/development/review-coordination-lobbi-docs-claude-2
SKILL.md
Review Coordination Skill
Coordinate multi-agent reviews, human reviews, and approval workflows for PRs.
When to Activate
This skill activates when:
- PR is ready for review
- User requests code review
- Agent council review is needed
- Review status needs checking
- Approval workflow needs management
Review Types
1. Agent Reviews
Automated reviews by specialized agents:
| Agent | Focus | Checks |
|---|---|---|
| code-agent | Code quality | Logic, patterns, performance |
| review-agent | Best practices | Standards, security, maintainability |
| doc-agent | Documentation | Comments, README, API docs |
| test-agent | Test coverage | Unit tests, integration tests |
| security-agent | Security | Vulnerabilities, secrets, auth |
2. Human Reviews
Team member reviews:
- Code owners
- Domain experts
- Tech leads
- Security team
3. Automated Checks
CI/CD status checks:
- Build status
- Test results
- Lint results
- Coverage thresholds
Review Workflow
Phase 1: Pre-Review Setup
-
Check PR Readiness:
mcp__harness__get_pullreq(pr_number)Verify:
- PR description complete
- All commits have proper messages
- No merge conflicts
- CI checks passing
-
Identify Reviewers:
- Code owners from changed paths
- Required reviewers from branch rules
- Suggested reviewers based on expertise
-
Assign Reviewers:
mcp__harness__add_reviewer({ pr_number: 45, reviewer: "alice" })
Phase 2: Agent Council Review
Invoke the agent council for comprehensive review:
Workflow:
┌─────────────────────────────────────────────────────────┐
│ Agent Council │
├─────────────────────────────────────────────────────────┤
│ ┌────────────┐ ┌────────────┐ ┌────────────┐ │
│ │ Code Agent │ │Review Agent│ │ Doc Agent │ │
│ └─────┬──────┘ └─────┬──────┘ └─────┬──────┘ │
│ │ │ │ │
│ ▼ ▼ ▼ │
│ ┌────────────────────────────────────────────────┐ │
│ │ Findings Aggregation │ │
│ └────────────────────────────────────────────────┘ │
│ │ │
│ ▼ │
│ ┌────────────────────────────────────────────────┐ │
│ │ Council Decision │ │
│ │ ✅ Approve | ⚠️ Request Changes | ❌ Block │ │
│ └────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────┘
Phase 3: Post Comments
For each finding, post structured comments:
## 🤖 Agent Review: Code Quality
**Agent:** code-agent
**Status:** ⚠️ Suggestions
### Findings
#### 1. Performance Concern
**File:** `src/auth/login.ts:45`
**Severity:** Medium
```typescript
// Current implementation
const user = await db.users.findAll().filter(u => u.email === email);
// Suggested improvement
const user = await db.users.findOne({ where: { email } });
Rationale: Using findAll followed by filter is inefficient. Use findOne with a where clause for better performance.
2. Missing Error Handling
File: src/auth/login.ts:52
Severity: Low
Consider adding try-catch for the token generation...
Summary
- 2 suggestions found
- 0 blocking issues
- Overall: Approve with suggestions
### Phase 4: Track Approvals
Monitor review status:
📋 PR #45 Review Status
Required Approvals: 2 Current Approvals: 1/2
Reviewers: ✅ @alice - Approved 🔄 @bob - Pending 💬 @charlie - Requested changes
Agent Reviews: ✅ code-agent - Approved with suggestions ✅ review-agent - Approved ✅ doc-agent - Approved
CI Checks: ✅ build - Passed ✅ test - Passed (98% coverage) ✅ lint - Passed
Status: Awaiting @bob's review and @charlie's re-review
### Phase 5: Sync to Jira
Update Jira issue with review status:
mcp__atlassian__jira_add_comment({ issue_key: "PROJ-123", body: `
PR Review Update
PR: #45 Status: In Review
Approvals
- ✅ @alice approved
- 🔄 @bob reviewing
- 💬 @charlie requested changes
Agent Council
All agents approved with minor suggestions.
Action Needed
- Address @charlie's feedback
- Await @bob's review ` })
## Approval Requirements
### Minimum Requirements
```yaml
review_requirements:
min_approvals: 2
required_reviewers:
- codeowner
dismiss_stale_reviews: true
require_agent_approval: true
require_ci_pass: true
Escalation Rules
| Condition | Action |
|---|---|
| PR open > 48h without review | Notify team lead |
| Blocking feedback > 24h | Notify assignee |
| All approvals received | Auto-notify for merge |
| Security finding | Block until addressed |
Review Templates
Approval Comment
✅ **Approved**
Reviewed the changes for:
- [x] Code quality
- [x] Performance
- [x] Security
- [x] Documentation
LGTM! Ready to merge.
Request Changes Comment
⚠️ **Changes Requested**
### Required Changes
1. Add error handling for edge case X
2. Update API documentation
### Suggestions
- Consider caching for better performance
Please address the required changes before merging.
Block Comment
❌ **Blocked**
### Critical Issues
1. **Security vulnerability detected**
- Potential SQL injection in `db.query()`
- Must use parameterized queries
This PR cannot be merged until the security issue is resolved.
cc: @security-team
Agent Review Commands
Trigger specific agent reviews:
# Full council review
/jira-harness:review PROJ-123 --council
# Specific agent review
/jira-harness:review PROJ-123 --agent code-agent
/jira-harness:review PROJ-123 --agent security-agent
# Re-review after changes
/jira-harness:review PROJ-123 --refresh
Best Practices
- Early Reviews: Request reviews early for complex changes
- Small PRs: Smaller PRs get faster, better reviews
- Clear Context: Provide context in PR description
- Address Feedback: Respond to all review comments
- Thank Reviewers: Acknowledge helpful feedback
- Learn from Reviews: Use feedback to improve
Didn't find tool you were looking for?