pr-review
Conduct comprehensive PR review from multiple perspectives (PM, Developer, QA, Security)
Best use case
pr-review is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
It is a strong fit for teams already working in Codex.
Conduct comprehensive PR review from multiple perspectives (PM, Developer, QA, Security)
Teams using pr-review should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/pr-review/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How pr-review Compares
| Feature / Agent | pr-review | Standard Approach |
|---|---|---|
| Platform Support | Codex | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Conduct comprehensive PR review from multiple perspectives (PM, Developer, QA, Security)
Which AI agents support this skill?
This skill is designed for Codex.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
Related Guides
Cursor vs Codex for AI Workflows
Compare Cursor and Codex for AI coding workflows, repository assistance, debugging, refactoring, and reusable developer skills.
AI Agents for Coding
Browse AI agent skills for coding, debugging, testing, refactoring, code review, and developer workflows across Claude, Cursor, and Codex.
AI Agents for Marketing
Discover AI agents for marketing workflows, from SEO and content production to campaign research, outreach, and analytics.
SKILL.md Source
# PR Review
**PR Link/Number**: $ARGUMENTS
> **Instructions**: Execute each task in the order given to conduct a thorough code review. Update GitHub with this review.
> **Important**: The future is now—any improvements or "future" recommendations must be addressed **immediately**.
---
## Arguments
- `<pr_link_or_number>` - GitHub PR URL or PR number (required)
- `--interactive` - Prompt for confirmation before posting review
- `--guidance "text"` - Additional review guidance or focus areas
- `--regression-gate` - Run regression check before merge approval (default: true)
- `--allow-regression` - Allow merge even if regressions detected (requires justification)
**Default behavior**: Regression gate is enabled by default. To disable: `--regression-gate=false`
---
## Task 1: Product Manager Review
**Objective**: Assess from a product management perspective, focusing on:
- **Business Value**: Does this PR clearly advance our core product goals and deliver immediate ROI?
- **User Experience**: Is the change intuitive and delightful for users right now? If not, make fixes immediately.
- **Strategic Alignment**: Does the PR align with current (and long-term, i.e., now) strategic objectives?
**Action**: Provide clear directives on how to ensure maximum user and business impact. All "future" suggestions must be implemented now.
---
## Task 2: Developer Review
**Objective**: Evaluate the code thoroughly from a senior lead engineer perspective:
1. **Code Quality & Maintainability**: Is the code structured for readability and easy maintenance? If not, refactor now.
2. **Performance & Scalability**: Will these changes operate efficiently at scale? If not, optimize immediately.
3. **Best Practices & Standards**: Note any deviation from coding standards and correct it now.
**Action**: Leave a concise yet complete review comment, ensuring all improvements happen immediately—no deferrals.
---
## Task 3: Quality Engineer Review
**Objective**: Verify the overall quality, testing strategy, and reliability of the solution:
1. **Test Coverage**: Are there sufficient tests (unit, integration, E2E)? If not, add them now.
2. **Potential Bugs & Edge Cases**: Have all edge cases been considered? If not, address them immediately.
3. **Regression Risk**: Confirm changes don't undermine existing functionality. If risk is identified, mitigate now with additional checks or tests.
**Action**: Provide a detailed QA assessment, insisting any "future" improvements be completed right away.
---
## Task 4: Security Engineer Review
**Objective**: Ensure robust security practices and compliance:
1. **Vulnerabilities**: Check for common security flaws (SQL injection, XSS, authentication bypass, etc.). Fix immediately if found.
2. **Sensitive Data**: Confirm no secrets, credentials, or PII are exposed. Remove immediately if found.
3. **Security Best Practices**: Verify adherence to OWASP guidelines and security standards. Address gaps now.
**Action**: Provide security assessment with immediate remediation for any findings. No deferring security issues.
---
## Task 5: Regression Detection Gate
**Objective**: Detect behavioral regressions introduced by this PR
**When**: Runs automatically before merge approval (unless `--regression-gate=false`)
**Process**:
1. **Identify base branch** from PR metadata
2. **Determine scope** using changed files
3. **Execute regression check**:
```bash
/regression-check \
--baseline <base-branch> \
--scope changed-files \
--format summary
```
4. **Analyze results**:
- **No regressions**: Proceed to merge approval
- **Minor regressions**: Flag for review, document in comment
- **Critical regressions**: BLOCK MERGE (unless `--allow-regression`)
**Output**: Regression analysis posted as PR comment
### Regression Gate Behavior
| Regression Severity | Default Action | With `--allow-regression` |
|---------------------|----------------|---------------------------|
| **None** | Approve merge | Approve merge |
| **Minor** (warnings) | Approve with note | Approve with note |
| **Major** (behavior change) | Request changes | Approve with documented risk |
| **Critical** (test failures) | BLOCK merge | Approve with justification required |
### Example: Regression Detected (Blocking)
```markdown
## 🚨 Regression Gate: BLOCKED
**Baseline**: main (commit abc123)
**Changed Files**: 5 files
**Tests Affected**: 12 tests
### Critical Regressions
1. **test/unit/auth/login.test.ts::validateCredentials**
- Status: PASS → FAIL
- Error: "Expected 200, received 401"
- Impact: Breaks user authentication
- **Action**: FIX REQUIRED BEFORE MERGE
2. **test/integration/api/payments.test.ts::processPayment**
- Status: PASS → FAIL
- Error: "Transaction timeout"
- Impact: Payment processing broken
- **Action**: FIX REQUIRED BEFORE MERGE
### Verdict
❌ **MERGE BLOCKED** - Critical regressions must be resolved.
To override (not recommended): Use `--allow-regression` and provide justification in PR description.
See full report: [regression-2026-01-25T15-30-00Z.md]
```
### Example: Regression Detected (Warning)
```markdown
## ⚠️ Regression Gate: WARNING
**Baseline**: main (commit abc123)
**Changed Files**: 3 files
**Tests Affected**: 8 tests
### Behavior Changes Detected
1. **test/integration/api/users.test.ts::createUser**
- Performance: 150ms → 450ms (+200%)
- Severity: MAJOR
- **Action**: INVESTIGATE before merge
### Verdict
⚠️ **MERGE WITH CAUTION** - Behavior changes detected but not blocking.
Recommendation: Review performance regression and document if intentional.
See full report: [regression-2026-01-25T15-30-00Z.md]
```
### Example: No Regressions
```markdown
## ✅ Regression Gate: PASSED
**Baseline**: main (commit abc123)
**Changed Files**: 2 files
**Tests Affected**: 5 tests
**Result**: All tests passing, no behavioral changes detected
### Verdict
✅ **REGRESSION CHECK PASSED** - Safe to merge from regression perspective.
```
**Integration with Regression Analyst**:
For deep regression analysis or complex regressions, escalate to @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/agents/regression-analyst.md for:
- Root cause analysis
- Regression pattern identification
- Fix recommendations
- Historical regression correlation
---
## Task 6: Review Summary
**Objective**: Synthesize all review feedback into actionable summary
**Required Sections**:
1. **Overall Verdict**: Approve / Request Changes / Comment
2. **Critical Issues**: Must be fixed before merge
3. **Major Issues**: Should be fixed before merge
4. **Minor Issues**: Can be addressed in follow-up
5. **Regression Status**: Results from regression gate
6. **Merge Recommendation**: Clear approve/block decision with rationale
### Example Summary
```markdown
## PR Review Summary
**Verdict**: REQUEST CHANGES
### Critical Issues (MUST FIX)
1. ❌ **Regression**: Authentication tests failing (see regression report)
2. ❌ **Security**: API key exposed in config file (line 42)
### Major Issues (SHOULD FIX)
1. ⚠️ **Performance**: Database query in loop (lines 156-178)
2. ⚠️ **Test Coverage**: New function missing unit tests
### Minor Issues (NICE TO HAVE)
1. 💡 **Code Style**: Inconsistent variable naming
2. 💡 **Documentation**: JSDoc comments missing
### Regression Status
❌ **2 CRITICAL REGRESSIONS DETECTED** - See detailed report above
### Merge Recommendation
🚫 **BLOCK MERGE**
**Rationale**:
- Critical regressions break authentication flow
- Security vulnerability exposes API credentials
- Must be resolved before merge
**Next Steps**:
1. Fix authentication regression
2. Remove exposed API key
3. Re-run regression check
4. Address major issues or create follow-up tickets
```
---
## Post-Review Actions
After completing review:
1. **Post review comment** to PR with summary
2. **Add labels** based on severity:
- `regression-risk` if regressions detected
- `security-issue` if vulnerabilities found
- `needs-tests` if coverage insufficient
- `performance-concern` if performance issues found
3. **Set PR status**:
- Approve (if all checks pass)
- Request Changes (if critical/major issues)
- Comment (if minor issues only)
4. **Save review artifacts**:
- Regression report (if gate enabled)
- Review summary
- Issue tracking links
---
## References
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/commands/regression-check.md - Regression detection command
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/agents/regression-analyst.md - Deep regression analysis agent
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/schemas/testing/regression.yaml - Regression detection schema
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/rules/executable-feedback.md - Executable feedback loop implementation
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/rules/executable-feedback.md - Test-first principlesRelated Skills
pr-reviewer
Review GitHub pull requests for code quality, security, and best practices. Use for automated PR feedback and approval workflows.
review-synthesis
Aggregate multi-reviewer feedback into consolidated, prioritized, and actionable recommendations
flow-security-review-cycle
Orchestrate continuous security validation, threat modeling, vulnerability management, and security gate enforcement across SDLC phases
budget-review
Project directory path (default current directory)
brand-review
Project directory path (default current directory)
aiwg-orchestrate
Route structured artifact work to AIWG workflows via MCP with zero parent context cost
venv-manager
Create, manage, and validate Python virtual environments. Use for project isolation and dependency management.
pytest-runner
Execute Python tests with pytest, supporting fixtures, markers, coverage, and parallel execution. Use for Python test automation.
vitest-runner
Execute JavaScript/TypeScript tests with Vitest, supporting coverage, watch mode, and parallel execution. Use for JS/TS test automation.
eslint-checker
Run ESLint for JavaScript/TypeScript code quality and style enforcement. Use for static analysis and auto-fixing.
repo-analyzer
Analyze GitHub repositories for structure, documentation, dependencies, and contribution patterns. Use for codebase understanding and health assessment.
YouTube Acquisition
yt-dlp patterns for acquiring content from YouTube and video platforms