SKILL.md
/test - Modular Project Audit
A comprehensive 20-phase autonomous project audit system with full GitHub integration.
Quick Reference
/test # Full audit (autonomous - fixes everything)
/test prodapp # Validate installed production app (Phase P)
/test docker # Validate Docker image and registry (Phase D)
/test github # Audit GitHub repository settings (Phase G)
/test --phase=A # Run single phase
/test --phase=0-2 # Run phase range
/test --phase=5,6 # Security + Dependencies only
/test --interactive # Enable interactive mode
/test help # Show helpKey Features
- Autonomous: Fixes ALL issues without prompting, loops until clean
- 20 Phases: Complete coverage from safety snapshots to documentation sync
- Multi-Language: Python, Node.js, Go, Rust, Shell, Docker, YAML
- 20+ Tools: ruff, pylint, bandit, trivy, CodeQL, and more
- GitHub Integration: Dependabot, CodeQL workflows, branch protection audit
- Context Efficient: 93% reduction via on-demand phase loading
Available Phases
| Phase | Name | Description |
|---|---|---|
| S | Snapshot | BTRFS safety snapshot |
| 0 | Pre-Flight | Environment validation + config audit + sandbox setup |
| 1 | Discovery | Detect project type, tools, GitHub |
| 2 | Execute | Run tests + analysis + coverage |
| 2a | Runtime | Service health checks |
| 5 | Security | CVE scanning, SAST |
| 6 | Dependencies | Package health |
| 7 | Quality | Linting, complexity, dead code cleanup |
| 10 | Fix | Auto-fix issues |
| A | App Test | Sandbox app testing |
| P | Production | Live app validation |
| D | Docker | Image validation |
| G | GitHub | Repo security audit |
| I | Infrastructure | Runtime issue detection |
| 12 | Verify | Re-run tests |
| 13 | Docs | Documentation sync |
| V | VM Testing | Heavy isolation testing |
| VM | Lifecycle | VM startup/shutdown management |
| C | Restore | Cleanup |
| ST | Self-Test | Framework self-validation (explicit only) |
Autonomous Behavior
The skill operates entirely non-interactively:
- Fix ALL Issues - No "manual required" lists, no deferrals, no exceptions
- Iterative Until Clean - After fixes are committed, the entire audit re-runs until a clean pass with zero issues
- All Audits Are Holistic - Cross-component analysis is a structural property of every analysis phase
- Documentation Sync - Docs always match codebase state
- Commit and Stage - All fixes committed; added to staged release if one exists
Tool Detection
Phase 1 automatically detects installed tools:
Code Quality: ruff, pylint, mypy, black, isort, eslint, prettier, hadolint, yamllint, shfmt, markdownlint, codespell
Security: pip-audit, bandit, npm audit, cargo audit, trivy, CodeQL
GitHub: gh CLI for repository auditing
Installation
Claude.ai (Web)
Upload this SKILL.md file directly to your Claude.ai account:
- Go to claude.ai → Settings → Skills
- Click "Upload skill" and select this file
- The skill reference will be available in your projects
Claude Code (CLI)
For full autonomous execution with all 20 phases:
# Clone and symlink
git clone https://github.com/TheBoscoClub/claude-test-skill.git ~/claude-test-skill
ln -s ~/claude-test-skill/commands/test.md ~/.claude/commands/test.md
ln -s ~/claude-test-skill/skills/test-phases ~/.claude/skills/test-phasesMore Information
- Repository: https://github.com/TheBoscoClub/claude-test-skill
- Full Documentation: See README.md and INSTALL.md
- Version: 4.1.1