crew-verification

Build a verification map wiring changed files to tests, credentials, and Playwright checks.

2|Updated Jul 28, 2026
One-click install
npx skills add https://github.com/mbadali25/useful-claude-add-ons --skill crew-verification-mbadali25
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: crew-verification
Source: https://github.com/mbadali25/useful-claude-add-ons/tree/main/plugin/crew/skills/crew-verification
Command: npx skills add https://github.com/mbadali25/useful-claude-add-ons --skill crew-verification-mbadali25

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams forget which tests should run after a change, and unmapped checks silently read as coverage. This Skill replaces agent memory with a deterministic .crew/verify.json map that a hook reads, so the decision of what to run is data-driven and reviewable. ## Core Features & Use Cases - Verification map authoring: Derive path-to-check rules from git history, verify each mapping by deliberately breaking code, and record the reasoning in a why field. - Safe credential handling: Record secret locations and retrieval commands in .crew/secrets.md without ever exposing values, preferring ephemeral containers and gitignored .env.smoke files over cloud secret stores. - Playwright and promotion gates: Set up tagged browser specs wired into the map, and define an environments block enforcing development-to-qa-to-production promotion with smoke, regression, and verify gates. - Use Case: After adding a database migration, the map automatically runs a fresh-database apply, a rollback test, and a round-trip check against a real database instead of relying on someone remembering to do it. ## Quick Start Ask the agent to build the verification map for this repository and wire the existing test directories into .crew/verify.json.

Frequently Asked Questions about crew-verification

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I decide which tests to run after a code change?

Build a verification map in .crew/verify.json that pairs file path patterns with the checks they require. Derive rules from git history of frequently changed files, then prove each mapping by breaking the code and confirming the test goes red.

How do I give tests credentials without exposing secrets?

Record only the secret's location and retrieval command in .crew/secrets.md, never the value. Prefer ephemeral containers with fixture data first, then a gitignored .env.smoke file, and a test-scoped secret store only as a last resort.

When should I use Playwright instead of unit or API tests?

Use Playwright only for what a browser can observe: rendering, layout, navigation, and full-stack integration. Business logic belongs in unit or API tests, which run far faster and do not break when UI elements move.

Why does a check that always passes fail to provide coverage?

A check that has never failed has never been shown capable of failing. Sabotage-test every check by breaking what it watches and confirming it goes red with a message naming the problem, then record the result.

What happens when a changed file matches no verification rule?

With "unmapped": "fail" set, the gate fails and names the unmatched file, converting forgotten test areas into blocking errors. Set it to "warn" only while first building the map.

Does the verification map enforce promotion gates outside Claude Code?

No. The promote-gate hook reads .crew/verify.json only inside a Claude Code session with the crew plugin active. The same deploy command runs unguarded in any session without the plugin, since enforcement lives in the hook, not the file.