GitHub Code Review Skill
Deploy specialized review agents (security, performance, architecture, style,
accessibility) to review PRs in parallel and post findings back through the
GitHub CLI — beyond traditional single-pass static analysis.
When to use
- Reviewing a pull request diff for security, performance, architecture, or style issues
- Running specialized checks (OWASP/secrets, Big-O/query efficiency, SOLID/coupling) on changed files
- Coordinating multiple review agents and posting inline comments, labels, or approve/request-changes decisions
- Wiring auto-review into CI (GitHub Actions) or PR-comment commands
When not to use
- Full project development with quality gates, TDD, and coverage analysis → use build-with-quality
- Release versioning, changelog generation, deployment orchestration → use github-release-management
- GitHub Actions workflow creation and CI/CD pipeline setup → use github-workflow-automation
- Cross-repository synchronisation and architecture management → use github-multi-repo
- Sprint planning, issue triage, project board management → use github-project-management
Quick start
# Initialize review swarm for PR
gh pr view 123 --json files,diff | npx ruv-swarm github review-init --pr 123
# Post review status
gh pr comment 123 --body "🔍 Multi-agent code review initiated"
Complete review workflow
# Get PR context with gh CLI
PR_DATA=$(gh pr view 123 --json files,additions,deletions,title,body)
PR_DIFF=$(gh pr diff 123)
# Initialize comprehensive review across specialized agents
npx ruv-swarm github review-init \
--pr 123 \
--pr-data "$PR_DATA" \
--diff "$PR_DIFF" \
--agents "security,performance,style,architecture,accessibility" \
--depth comprehensive
Claude Code only: the swarm orchestration above (ruv-swarm/claude-flow) needs
the claude-flow MCP server. On Codex / GPT-6 Astra (or any session without it): see
references/plain-gh-fallback.md for the plain
gh-only path.
Reference library
Detailed agent catalogs, config schemas, and end-to-end recipes live in references/.
Load the one you need on demand:
- references/review-agents.md — the specialized
agents (security, performance, architecture, style), the checks each performs,
comment templates, and how to build/register a custom review agent.
- references/swarm-and-automation.md —
PR-driven swarm creation, label→agent mapping, topology-by-PR-size, PR-comment
commands, webhook handler, GitHub Actions auto-review, inline comment generation,
CI/CD integration, auto-fix and auto-merge.
- references/configuration-and-gates.md —
.github/review-swarm.yml, custom path triggers, quality-gate thresholds, status
checks, security considerations/checklist, and review/comment best practices.
- references/advanced-and-workflows.md —
context-aware and cross-PR reviews, learning-from-history, five complete workflow
examples (security/performance/UI/feature/bug-fix), monitoring/analytics, Claude
Code integration, and troubleshooting.
Related skills
github-release-management — version management and deployment orchestrationgithub-workflow-automation — CI/CD pipeline managementgithub-project-management — issue tracking and sprint planninggithub-multi-repo — cross-repository coordination
Last Updated: 2026-07-28 · Version: 1.0.0 · Maintainer: Claude Code Flow Team ·
Licensed under the MIT License as part of the Claude Code Flow project.
1---2name: github-code-review3description: Multi-agent code review for GitHub PRs. Use when reviewing a pull request, running security/performance/architecture/style checks on a diff, or coordinating specialized review agents and posting findings back to a PR.4---56# GitHub Code Review Skill78Deploy specialized review agents (security, performance, architecture, style,9accessibility) to review PRs in parallel and post findings back through the10GitHub CLI — beyond traditional single-pass static analysis.1112## When to use1314- Reviewing a pull request diff for security, performance, architecture, or style issues15- Running specialized checks (OWASP/secrets, Big-O/query efficiency, SOLID/coupling) on changed files16- Coordinating multiple review agents and posting inline comments, labels, or approve/request-changes decisions17- Wiring auto-review into CI (GitHub Actions) or PR-comment commands1819## When not to use2021- Full project development with quality gates, TDD, and coverage analysis → use **build-with-quality**22- Release versioning, changelog generation, deployment orchestration → use **github-release-management**23- GitHub Actions workflow creation and CI/CD pipeline setup → use **github-workflow-automation**24- Cross-repository synchronisation and architecture management → use **github-multi-repo**25- Sprint planning, issue triage, project board management → use **github-project-management**2627## Quick start2829```bash30# Initialize review swarm for PR31gh pr view 123 --json files,diff | npx ruv-swarm github review-init --pr 1233233# Post review status34gh pr comment 123 --body "🔍 Multi-agent code review initiated"35```3637### Complete review workflow3839```bash40# Get PR context with gh CLI41PR_DATA=$(gh pr view 123 --json files,additions,deletions,title,body)42PR_DIFF=$(gh pr diff 123)4344# Initialize comprehensive review across specialized agents45npx ruv-swarm github review-init \46 --pr 123 \47 --pr-data "$PR_DATA" \48 --diff "$PR_DIFF" \49 --agents "security,performance,style,architecture,accessibility" \50 --depth comprehensive51```5253**Claude Code only:** the swarm orchestration above (`ruv-swarm`/`claude-flow`) needs54the claude-flow MCP server. On Codex / GPT-6 Astra (or any session without it): see55[references/plain-gh-fallback.md](references/plain-gh-fallback.md) for the plain56`gh`-only path.5758## Reference library5960Detailed agent catalogs, config schemas, and end-to-end recipes live in `references/`.61Load the one you need on demand:6263- **[references/review-agents.md](references/review-agents.md)** — the specialized64 agents (security, performance, architecture, style), the checks each performs,65 comment templates, and how to build/register a custom review agent.66- **[references/swarm-and-automation.md](references/swarm-and-automation.md)** —67 PR-driven swarm creation, label→agent mapping, topology-by-PR-size, PR-comment68 commands, webhook handler, GitHub Actions auto-review, inline comment generation,69 CI/CD integration, auto-fix and auto-merge.70- **[references/configuration-and-gates.md](references/configuration-and-gates.md)** —71 `.github/review-swarm.yml`, custom path triggers, quality-gate thresholds, status72 checks, security considerations/checklist, and review/comment best practices.73- **[references/advanced-and-workflows.md](references/advanced-and-workflows.md)** —74 context-aware and cross-PR reviews, learning-from-history, five complete workflow75 examples (security/performance/UI/feature/bug-fix), monitoring/analytics, Claude76 Code integration, and troubleshooting.7778## Related skills7980- `github-release-management` — version management and deployment orchestration81- `github-workflow-automation` — CI/CD pipeline management82- `github-project-management` — issue tracking and sprint planning83- `github-multi-repo` — cross-repository coordination8485---8687**Last Updated:** 2026-07-28 · **Version:** 1.0.0 · **Maintainer:** Claude Code Flow Team ·88Licensed under the MIT License as part of the Claude Code Flow project.
dreamlab-ai