By fzymgc-house
PR review workflow with specialized review agents and automated fix loop
npx claudepluginhub fzymgc-house/fzymgc-house-skills --plugin pr-reviewDetects breaking changes, backward compatibility issues, and contract violations in PRs. Used by the review-pr orchestrator for the `api` aspect.
Reviews code for project guideline compliance, bugs, and quality issues. Used by the review-pr orchestrator for the `code` aspect.
Improves code clarity, consistency, and maintainability while preserving functionality. Used by the review-pr orchestrator for the `simplify` aspect.
Analyzes code comments for accuracy, completeness, and long-term maintainability. Used by the review-pr orchestrator for the `comments` aspect.
Implements code fixes for PR review findings. Used by the address-findings orchestrator. Receives a finding bead ID, reads the finding details, and applies the fix in its isolated worktree.
Analyzes test coverage quality and identifies critical testing gaps in PRs. Used by the review-pr orchestrator for the `tests` aspect.
Validates that code fixes correctly address their review findings. Used by the address-findings orchestrator after fix commits are integrated. Receives finding IDs and a VCS diff, returns PASS/FAIL per finding.
Audits PR changes for security vulnerabilities using OWASP methodology. Used by the review-pr orchestrator for the `security` aspect.
Audits error handling for silent failures, inadequate handlers, and inappropriate fallbacks. Used by the review-pr orchestrator for the `errors` aspect.
Verifies implementation alignment with design documents, ADRs, and architectural decisions. Used by the review-pr orchestrator for the `spec` aspect.
Analyzes type designs for invariant strength, encapsulation quality, and practical usefulness. Used by the review-pr orchestrator for the `types` aspect.
Validates fixes and runs project quality gates after fixes are applied. Used by the address-findings orchestrator. Receives a fix manifest with problem/fix/change context and verifies alignment plus lint/build/tests.
Processes findings from review-pr by working through beads in the review epic. Use when the user asks to "address review findings", "fix review issues", "work through review beads", or "process review findings".
Manages GitHub PR review comments. Use when the user asks to "list PR comments", "check PR feedback", "respond to PR comments", "address reviewer feedback", "fix PR review issues", "work through PR review", or "acknowledge comments". Handles both quick lookups and full review-response workflows.
Comprehensive PR review using specialized agents. Use when the user asks to "review this PR", "check my PR", "review code quality", "run PR review", "analyze this pull request", or invokes /review-pr. Launches targeted review agents for code quality, error handling, test coverage, type design, comments, security, API compatibility, spec compliance, and code simplification.
A Claude Code plugin marketplace for the fzymgc-house self-hosted cluster. Contains two plugins covering homelab infrastructure operations and automated PR review workflows.
Infrastructure skills for interacting with the homelab cluster.
| Skill | Description |
|---|---|
| grafana | Grafana, Loki, and Prometheus operations — dashboards, logs, metrics, alerting, incidents, OnCall, profiling |
| terraform | Terraform Cloud operations — runs, workspaces, state management, registry documentation |
| skill-qa | Validates SKILL.md files against Claude Code best practices |
Automated PR review workflow using 12 specialized agents with git worktree isolation for parallel execution.
Skills (orchestrators):
| Skill | Description |
|---|---|
| review-pr | Dispatches up to 9 review agents in parallel, persists findings as beads |
| address-findings | Fix loop with worktree-isolated fix-workers, merge protocol, and review gates |
| respond-to-comments | GitHub PR comment management with bead-aware context |
Agents (dispatched by orchestrator skills):
| Agent | Role |
|---|---|
| code-reviewer | Project guideline compliance and bug detection |
| silent-failure-hunter | Error handling and silent failure auditing |
| pr-test-analyzer | Test coverage quality and gap analysis |
| type-design-analyzer | Type invariant strength and encapsulation |
| comment-analyzer | Code comment accuracy and maintainability |
| security-auditor | OWASP-based security vulnerability detection |
| api-contract-checker | Breaking changes and backward compatibility |
| spec-compliance | Alignment with design docs and ADRs |
| code-simplifier | Clarity and maintainability improvements |
| fix-worker | Implements fixes in isolated worktrees |
| review-gate | Validates fixes address their findings |
| verification-runner | Runs quality gates after fixes are applied |
claude plugin install github:fzymgc-house/fzymgc-house-skills
Or install individual plugins:
claude plugin install github:fzymgc-house/fzymgc-house-skills/homelab
claude plugin install github:fzymgc-house/fzymgc-house-skills/pr-review
git clone git@github.com:fzymgc-house/fzymgc-house-skills.git
cd fzymgc-house-skills
lefthook install
.claude-plugin/
marketplace.json # Marketplace manifest listing both plugins
homelab/
plugin.json # Plugin manifest
skills/
grafana/ # Grafana/Loki/Prometheus operations
terraform/ # Terraform Cloud operations
skill-qa/ # SKILL.md validation
pr-review/
plugin.json # Plugin manifest
agents/ # 12 agent definitions (YAML frontmatter + system prompt)
skills/
review-pr/ # Review orchestrator
address-findings/ # Fix loop orchestrator
respond-to-comments/ # PR comment management
All commits must follow Conventional Commits format. This is enforced by a commit-msg hook via cocogitto.
feat(grafana): add incident management support
fix(review-pr): correct agent dispatch for security aspect
docs: update README
Versions are managed by release-please. Do not manually bump versions in marketplace.json, plugin.json, or SKILL.md files.
Private repository for fzymgc-house infrastructure.
Team-oriented workflow plugin with role agents, 27 specialist agents, ECC-inspired commands, layered rules, and hooks skeleton.
Uses power tools
Uses Bash, Write, or Edit tools
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Comprehensive startup business analysis with market sizing (TAM/SAM/SOM), financial modeling, team planning, and strategic research
Complete collection of battle-tested Claude Code configs from an Anthropic hackathon winner - agents, skills, hooks, rules, and legacy command shims evolved over 10+ months of intensive daily use