Prunes redundant tests and detects critical coverage gaps through multi-phase test suite auditing.
npx skills add https://github.com/mcinnisd/gymbro --skill deep-test-mcinnisd
Please help me install this Agent Skill. Skill: deep-test Source: https://github.com/mcinnisd/gymbro/tree/main/.agents/skills/deep-test Command: npx skills add https://github.com/mcinnisd/gymbro --skill deep-test-mcinnisd
gh issue create. - Use Case: Run it before a release to consolidate a slow test suite, verify auth isolation across endpoints, and automatically ticket remaining gaps for implementation. ## Quick Start Ask the agent to run a deep test audit that prunes redundant tests, executes the suite, and files issues for any critical coverage gaps it finds.High-intent search queries and answers about installing and using this skill.
Delete tautological tests that only assert mocks were called, fake E2E tests that are fully mocked, and duplicate flow tests covering identical endpoints. Consolidate repeated sequences into parameterized fixtures while keeping tests that target public domain boundaries.
Audit public seams rather than internals: check multi-tenant isolation across all endpoints, negative auth paths like expired JWTs and duplicate registrations, and full round-trips from API payload to database mutation. Run pytest with MOCK_DB enabled to verify behavior.
A tautological test mocks the function under test and merely asserts the mock was called, verifying nothing about real behavior. Tests asserting internal keys of mock classes or static dictionary strings also fall into this category.
Yes, the workflow pauses after presenting a test health and net reduction report. GitHub issues are only created via gh issue create after the human confirms the proposed deletions, fixes, and tickets.
Keep mocked tests when they verify realistic failure paths such as rate limits, timeouts, or malformed inputs that are hard to reproduce live. Replace them only when a real vertical tracer-bullet test can cover the same public seam.
compozyGuides test placement by identifying invariants, owning layers, and canonical suites before changing coverage.
mst-software-vnImplements work described in a PRD or issues using TDD and code review.
swyxioAudit and harden repository test suites by pruning low-value tests and adding regression coverage.
czer323Generate and analyze tests across unit, property-based, mutation, and contract workflows.