qa-plan

Audit TDD plan documents with a 12-point adversarial gate and emit PASS or HOLD verdicts.

Updated Mar 22, 2026
One-click install
npx skills add https://github.com/pollockjj/grader-in-the-loop --skill qa-plan-pollockjj
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: qa-plan
Source: https://github.com/pollockjj/grader-in-the-loop/tree/main/.claude/skills/qa-plan
Command: npx skills add https://github.com/pollockjj/grader-in-the-loop --skill qa-plan-pollockjj

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Adversarial gatekeeping for TDD plan documents ensures plans are diagnosable, artifact-backed, and reviewable before implementation, reducing risk of flawed plans.

Core Features & Use Cases

  • Enforces a fixed 12-point audit on plan documents to guarantee diagnosability and verifiability.
  • Produces deterministic PASS or HOLD verdicts with per-check findings and no partial credit.
  • Prevents plan review bias by enforcing strict provenance, artifact requirements, and objective criteria.

Quick Start

Inspect a TDD plan with qa-plan to obtain a PASS or HOLD verdict and a detailed per-check report.

Frequently Asked Questions about qa-plan

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

FAQPage Schema
How do I audit a TDD plan document before implementation?

The audit applies a 12-point adversarial review to a TDD plan document or GitHub issue body, checking diagnosability and artifact backing, then emits a PASS or HOLD verdict with detailed per-check findings.

What is adversarial gatekeeping for TDD plans?

Adversarial gatekeeping for TDD plans is a strict review process that prevents plan review bias by enforcing objective criteria and artifact requirements, ensuring plans are diagnosable and reviewable before implementation begins.

How do I enforce artifact requirements in a TDD plan review?

You enforce artifact requirements in a TDD plan review by applying an adversarial audit that checks strict provenance and artifact backing, producing deterministic PASS or HOLD verdicts with no partial credit.

Can I use a GitHub issue body as input for a TDD plan audit?

Yes, you can use a GitHub issue body as input for a TDD plan audit, as the review process reads plan documents or GitHub issue bodies and applies its 12-point audit to emit a verdict.

Why does my TDD plan audit return a HOLD verdict instead of PASS?

Your TDD plan audit returns a HOLD verdict because the review process is HOLD-default, meaning any failure across the 12-point adversarial checks will prevent a PASS until all findings are resolved.

Does the TDD plan audit give partial credit for incomplete plan sections?

No, the TDD plan audit does not give partial credit; it produces deterministic PASS or HOLD verdicts based on strict objective criteria across all 12 adversarial checks.