Quality

Harness Audit checks a Claude project’s CLAUDE.md, hooks, skills, and knowledge files against seven health principles. Use it to find missing coverage, stale docs, and rule gaps.

Internal Review runs an L1 checklist for code, specs, designs, functions, and document integrity, then records the result in the development plan. Use it to verify recent work and decide the next Gate step.

OWASP Security Audit checks security-sensitive code against the OWASP Top 10:2025. Use it when reviewing authentication, input handling, APIs, data storage, or outbound requests.

Penetration Testing Checklist provides a structured workflow for authorized web application security testing, including recon, scanning, exploitation validation, and reporting. Use it for pentests, bug bounties, or pre-deployment security checks.

Playwright E2E Testing builds end-to-end tests for critical user flows, regression coverage, and integration checks. It uses Feature Object patterns, cross-browser runs, visual regression, and network mocking.

PM Gate Review checks a gate submission against a 6-item checklist and produces a recommendation. Use it when evaluating whether a G2, G3, or G4 submission is complete and compliant.

Snyk Agent Scan Compliance fixes snyk-agent-scan alerts in Claude Code skill files by restructuring content. It is used when authoring or triaging a skill that needs W001, W011, or W012 remediation.

SOP Review guides an L1 code review by loading the task, checking standards, reviewing the diff, and making an approve or reject decision. Use it when you need a structured review that must not skip checkpoints.

TDD Workflow enforces red-green-refactor development with tests written before implementation. Use it for new features, bug fixes, and refactors when you want behavior verified by failing tests first.

Testing Strategy applies RED-GREEN-REFACTOR TDD, separates unit and integration tests, and sets coverage targets for different feature types. Use it when writing tests, reviewing coverage, or testing database and admin flows.

Agent Integration Testing guides agents to write verifiable integration test specs and execute them through subagents. Use it when validating features, checking failures, or documenting test plans with reproducible steps and expectations.

Bugbash systematically explores software projects to find bugs, usability issues, and edge cases. It is used on CLIs, APIs, backends, and libraries to produce reproducible reports with commands, logs, and traces.