test-audit · fcakyon — Claude Code skill: test audits and setup
The test-audit skill reviews new and changed tests, and helps spot duplicates or low-value checks. Use it when adding tests or auditing an existing suite.
MIT ★ 1,216 updated 2026-10-24
/plugin marketplace add fcakyon/claude-codex-settings
/plugin install test-audit@claude-settingsWhat test-audit does
Before a test is added, the skill asks what observable contract or invariant it protects, which regression would make it fail, and why current tests miss that case. Each contract should have a primary test at its strongest boundary. A second layer needs a separate transport or lifecycle risk. Checks tied to implementation details, repeated contracts, or test-only seams without production callers are candidates for rejection.
For an audit, inspect the full owners and callers, change history, and the proof that will remain before deleting a candidate. Start a whole-subsystem campaign by reading references/campaign.md; candidate criteria are in references/audit.md. After edits, run owner and sibling tests, perform required checks, and review the full diff. The audit weighs the evidence lost by a deletion, rather than aiming for a particular number of removed tests.
What it helps with
- check which regression a new test would catch
- find tests that repeat the same contract
- reject checks coupled to implementation details
- assess remaining proof before removing a test
- review test-only seams with no production caller
Example prompts for the agent
- Review these tests for duplicates and low-value checks.
- Assess whether this change needs a test and what regression it should catch.
- Audit this subsystem's tests before removing candidates.
How to install test-audit
1Plugin marketplace
/plugin marketplace add fcakyon/claude-codex-settings/plugin install test-audit@claude-settingsThis installs the whole test-audit plugin with all its skills
2Skills CLI (Vercel)
npx skills add fcakyon/claude-codex-settings --skill test-audit -a claude-code3Manually: put the files in place
~/.claude/skills/test-audit/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/test-audit/
1Skills CLI (Vercel)
npx skills add fcakyon/claude-codex-settings --skill test-audit -a codex2Manually: put the files in place
.agents/skills/test-audit/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add fcakyon/claude-codex-settings --skill test-audit -a cursor2Manually: put the files in place
.cursor/skills/test-audit/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install fcakyon/claude-codex-settings plugins/test-audit/skills/test-audit2Manually: put the files in place
.github/skills/test-audit/SKILL.md1Skills CLI (Vercel)
npx skills add fcakyon/claude-codex-settings --skill test-audit -a gemini-cli2Platform CLI
gemini skills install https://github.com/fcakyon/claude-codex-settings --path plugins/test-audit/skills/test-auditManually: put the files in place
.devin/skills/test-audit/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about test-audit
How do I install test-audit in Claude Code?
Add the marketplace with /plugin marketplace add fcakyon/claude-codex-settings, then install the skill with /plugin install test-audit@claude-settings.
What does the test-audit skill do?
It evaluates new tests and helps identify duplicate, low-value, or implementation-coupled checks. During an audit, it considers which contracts remain covered after a test is removed.
What should I read before auditing a whole subsystem?
Begin with references/campaign.md. Before removing candidates, inspect their owners and callers, history, and the proof that will remain.
Source text
SKILL.md · MIT
--- name: test-audit description: This skill should be used when the user asks to "write a test", "review these tests", "audit tests", "prune low-value tests", or "find duplicate tests", and whenever writing, changing, reviewing, or sweeping tests. Gates new tests and audits low-value, implementation-coupled, or duplicative tests and the test-only production seams they demand. --- # Test Audit - Authoring gates every new/changed test. - Audits: remove low-value proof in coherent owner batches; never chase deletion counts. - Whole-subsystem campaign: read [references/campaign.md](references/campaign.md) first. ## Authoring gate Before adding test, answer: 1. Protected observable behavior, invariant, or independent contract? 2. Credible regression that makes it fail? 3. Why existing proof misses it? Each contract: one primary owner at strongest boundary. Another layer needs distinct transport/lifecycle risk. Extend tables/shared fixtures; consolidate repeated setup. 4. Export, flag, wrapper, injection hook with no production caller needed? Move proof to real boundary instead. - Use the repository testing guide for boundary and cost decisions. - Regressions must fail pre-fix for intended reason, pass after repair; report unavailable baseline proof. - One owner regression covers bug; never replay at every crossed layer. - Behavior-preserving refactors should not break test. ## Junk patterns - Reject self-comparisons, copied inventories, source greps, implementation/call-shape assertions, duplicate contracts, test-only seams, subject-generated expected values, mocks supplying asserted behavior. - Fixtures must not fabricate owner receipts/order or claim persistence path never writes. - Negative controls must fail at intended guard; names must match exercised behavior. - Auditing candidates: use [full checklist](references/audit.md#junk-patterns). ## Retention bar - Keep independent API, SDK, protocol, config, migration, storage, security, platform, default, prompt-byte, generated-language, package, release, architecture, observable-ordering contracts. Static/slow ≠ disposable. - Source inspection must survive identifier-only refactors and guard independent user-facing contract. - Baseline failures may be bugs to repair, not tests to delete. ## Audit and validation - Before audit edit: read [candidate evidence and workflow](references/audit.md#candidate-evidence); inspect full owners/callers/history; record remaining proof before deletion. - Delete obsolete production seams with tests; never optimize for deletion counts. - Follow the repository testing guide; preserve active test checkouts; run owner and sibling tests plus required checks, then review the full diff. - Report causes, retained contracts, proof/cost, production versus test LOC, follow-ups. Commit/push/land only with task authority.
© fcakyon · MIT · fcakyon/claude-codex-settings
Similar skills
All skills for Testing and QA →test-audit by other authors comes first.