From glincker-claude-code-marketplace
Autonomous AI-powered pull request reviewer with multi-agent analysis and comprehensive feedback
npx claudepluginhub joshuarweaver/cascade-code-general-misc-4 --plugin glincker-claude-code-marketplaceThis skill is limited to using the following tools:
**⚡ UNIQUE FEATURE**: Multi-agent review system with specialized reviewers for security, performance, testing, and architecture - the first autonomous PR review skill with parallel agent coordination.
Generates design tokens/docs from CSS/Tailwind/styled-components codebases, audits visual consistency across 10 dimensions, detects AI slop in UI.
Records polished WebM UI demo videos of web apps using Playwright with cursor overlay, natural pacing, and three-phase scripting. Activates for demo, walkthrough, screen recording, or tutorial requests.
Delivers idiomatic Kotlin patterns for null safety, immutability, sealed classes, coroutines, Flows, extensions, DSL builders, and Gradle DSL. Use when writing, reviewing, refactoring, or designing Kotlin code.
⚡ UNIQUE FEATURE: Multi-agent review system with specialized reviewers for security, performance, testing, and architecture - the first autonomous PR review skill with parallel agent coordination.
Automatically reviews pull requests with multiple specialized AI agents working in parallel:
Unlike simple code review tools, this skill:
Identify the PR:
git diff main...HEADgh pr view <number> to get PR detailsGather Context:
Launch 5 specialized Task agents in parallel:
Agent 1: Security Reviewer
Task: Security analysis of PR
Prompt: "Analyze these code changes for security vulnerabilities:
- SQL injection risks
- XSS vulnerabilities
- Hardcoded secrets or API keys
- Authentication/authorization issues
- Dependency vulnerabilities
- OWASP Top 10 issues
Files: [list changed files]
Provide:
1. Severity ratings (Critical/High/Medium/Low)
2. Specific line numbers
3. Exploitation scenarios
4. Remediation steps with code examples"
Agent 2: Performance Reviewer
Task: Performance analysis of PR
Prompt: "Analyze these code changes for performance issues:
- Inefficient algorithms (O(n²) vs O(n log n))
- Database N+1 queries
- Memory leaks
- Unnecessary re-renders (React/Vue)
- Blocking operations
- Resource waste
Files: [list changed files]
Provide:
1. Performance impact assessment
2. Specific bottlenecks with line numbers
3. Benchmark comparison suggestions
4. Optimized code examples"
Agent 3: Testing Reviewer
Task: Test coverage analysis of PR
Prompt: "Analyze test coverage and quality:
- Calculate test coverage for changed code
- Identify untested edge cases
- Review test quality and assertions
- Suggest additional test scenarios
- Check for test best practices
Files: [list changed files]
Provide:
1. Coverage percentage
2. Missing test cases
3. Test improvement suggestions
4. Example test code"
Agent 4: Architecture Reviewer
Task: Architecture and design analysis
Prompt: "Review architectural decisions:
- Design pattern appropriateness
- SOLID principles adherence
- Code modularity and coupling
- Separation of concerns
- Scalability considerations
- Technical debt introduced
Files: [list changed files]
Provide:
1. Architecture assessment
2. Design improvement suggestions
3. Refactoring recommendations
4. Long-term impact analysis"
Agent 5: Style & Documentation Reviewer
Task: Code style and documentation review
Prompt: "Review code style and documentation:
- Naming conventions
- Code readability
- Comment quality
- API documentation
- README updates needed
- Breaking changes documented
Files: [list changed files]
Provide:
1. Style issues with line numbers
2. Documentation gaps
3. Readability improvements
4. Suggested comments"
Collect all agent results (wait for all Task agents to complete)
Generate comprehensive review:
# PR Review Summary
## 📊 Overview
- Files changed: X
- Lines added: Y
- Lines removed: Z
- Overall rating: [Excellent/Good/Needs Work/Reject]
## 🔒 Security (Critical: X, High: Y, Medium: Z)
[Agent 1 findings summary]
## ⚡ Performance (Issues: X)
[Agent 2 findings summary]
## ✅ Testing (Coverage: X%)
[Agent 3 findings summary]
## 🏗️ Architecture
[Agent 4 findings summary]
## 📝 Style & Documentation
[Agent 5 findings summary]
## 🎯 Action Items
1. [Priority action with fix]
2. [Priority action with fix]
## 💡 Suggested Changes
[Code blocks with suggested improvements]
## ✨ Highlights
[Positive aspects of the PR]
Generate fix suggestions:
pr-review-fixes.md file with all suggested changespr-review-fixes.patch fileOffer the user:
gh pr comment <number> -F pr-review.mdUser Request: "Review PR #123"
Workflow:
gh pr view 123Output:
🔍 PR #123 Review Complete
📊 Overall: Good (minor improvements needed)
🔒 Security: ✅ No issues found
⚡ Performance: ⚠️ 1 issue found
✅ Testing: ⚠️ Coverage 78% (target: 80%)
🏗️ Architecture: ✅ Well designed
📝 Style: ⚠️ 3 minor issues
📋 Action Items:
1. Add database index for user_id column (performance)
2. Add tests for error scenarios (testing)
3. Update function documentation (style)
Would you like me to:
1. Post this review as a PR comment
2. Apply the suggested fixes
3. Generate the missing tests
User Request: "Review my current changes before I push"
Workflow:
git diff main...HEADCustomize review behavior:
# .pr-reviewer-config.yml
agents:
security:
enabled: true
severity_threshold: medium
performance:
enabled: true
benchmark_required: false
testing:
enabled: true
min_coverage: 80
architecture:
enabled: true
check_solid: true
style:
enabled: true
follow_existing: true
review:
auto_post_comment: false
suggest_fixes: true
blocking_issues: [critical_security, zero_tests]
gh CLI installed for PR operationsReview only new commits since last review:
git diff PR_BASE...HEAD --since="last review"
Add your own specialized agents:
Learns from approved/rejected reviews to adapt to team preferences.
Can integrate with:
This is a flagship skill for GLINCKER Marketplace. Contributions welcome:
Apache License 2.0 - See LICENSE
GLINCKER Team
🌟 This is a UNIQUE skill not available in other marketplaces!