test-evidence-review

Reviews test files and manual evidence documents for assertion coverage, sign-offs, and completeness.

13|Updated Jul 30, 2026
One-click install
npx skills add https://github.com/frabcd/codex-ai-game-studio --skill test-evidence-review-frabcd
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: test-evidence-review
Source: https://github.com/frabcd/codex-ai-game-studio/tree/main/plugins/ai-game-studio/skills/test-evidence-review
Command: npx skills add https://github.com/frabcd/codex-ai-game-studio --skill test-evidence-review-frabcd

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? A test file that exists and passes may still leave critical behaviour uncovered, and a manual evidence document may lack required sign-offs. This Skill audits the quality of test files and evidence documents for game production stories, assigning an ADEQUATE, INCOMPLETE, or MISSING verdict per story before QA sign-off. ## Core Features & Use Cases - Automated test quality review: Evaluates assertion coverage per test function, edge case handling, naming conventions, and formula traceability for Logic and Integration stories. - Manual evidence review: Checks criterion linkage, developer/designer/QA sign-off completeness, screenshot artefacts, and evidence freshness for Visual/Feel and UI stories. - Verdict reporting: Produces a story-by-story report with BLOCKING and ADVISORY issues, plus an overall sprint or system verdict, optionally written to production/qa/evidence-review-[date].md. - Use Case: Before closing a sprint, run the review across all stories to find a Logic story whose test passes vacuously with zero assertions and a Visual/Feel story missing the QA lead sign-off, then resolve both before story-done. ## Quick Start Ask the assistant to run a test evidence review on the current sprint and report which stories have adequate, incomplete, or missing test evidence.

Frequently Asked Questions about test-evidence-review

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

FAQPage Schema
How do I review test quality before QA sign-off?

Run the review scoped to a single story, the current sprint, or a system. It reads each story's Test Evidence section, locates the test or evidence file, and evaluates assertion coverage, edge cases, naming, and sign-offs, returning an ADEQUATE, INCOMPLETE, or MISSING verdict per story.

What makes a test file fail a quality review?

A test with zero assertions is flagged as BLOCKING because it passes vacuously. One to two assertions per function is noted as thin, generic names like test_1 are flagged as naming issues, and missing edge case or formula references reduce the verdict to INCOMPLETE.

Does this skill modify or fix test files?

No. It only reads and evaluates test files and evidence documents, then reports quality issues. Fixing BLOCKING or ADVISORY items is left to the user, optionally with scaffolded assertion patterns from a separate test-helpers skill.

What sign-offs are required in manual evidence documents?

Evidence documents for Visual/Feel and UI stories should contain developer, designer or art-lead, and QA lead sign-off lines. Missing or blank sign-offs flag the story as INCOMPLETE, since it cannot be fully closed without all required approvals.

When should I run a test evidence review?

Run it before QA hand-off sign-off, on any story whose test quality is in question, or during milestone reviews of Logic and Integration stories. It complements smoke checks that only verify tests exist and pass.