deep-qa

Identify and prioritize defects in software artifacts before implementation.

4|Updated Feb 23, 2026
One-click install
npx skills add https://github.com/npow/claude-skills --skill deep-qa
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deep-qa
Source: https://github.com/npow/claude-skills/tree/main/deep-qa
Command: npx skills add https://github.com/npow/claude-skills --skill deep-qa

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Deep QA identifies defects, contradictions, omissions, security risks, and unverified assumptions in existing artifacts before they reach implementation or production.

Core Features & Use Cases

  • Adversarial Critique: Reviews specifications, code, diffs, pull requests, research reports, skills, prompts, and technical documents through independent QA dimensions.
  • Structured Defect Detection: Coordinates parallel critics, deduplicates findings, validates severity with independent judges, and produces an actionable defect registry.
  • Verification Workflows: Runs tests, type checks, builds, smoke tests, and evidence-gathering workflows when verifying whether a change actually works.
  • Use Case: Give the skill a recent git diff to identify missed contract updates, unsafe error paths, ordering problems, resource exhaustion risks, and test coverage gaps before merging.

Quick Start

Use the deep-qa skill to review the artifact at the specified path and report all verified defects with severity, scenarios, root causes, and remediation directions.

Frequently Asked Questions about deep-qa

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

FAQPage Schema
How do I find defects in a git diff before merging?

Adversarial code review detects defects in a git diff by coordinating parallel critic agents to identify missed contract updates, unsafe error paths, and resource exhaustion risks before merging. This structured verification produces an actionable registry of prioritized defects.

What is adversarial specification audit for technical documents?

Adversarial specification audit is a verification process that identifies contradictions, omissions, security risks, and unverified assumptions in technical documents before implementation. It applies independent QA dimensions and structured defect detection to ensure artifact completeness and consistency.

Can I run smoke tests and type checks to verify a code change actually works?

Yes, you can run verification workflows that execute tests, type checks, builds, and smoke tests to gather evidence verifying whether a code change actually works. These workflows integrate with the adversarial critique to validate functional correctness alongside defect detection.

How do I prioritize security risks and omissions found during code review?

You prioritize security risks and omissions found during code review by validating their severity with independent judges. The system deduplicates findings from parallel critics and produces a structured defect registry that includes severity, scenarios, root causes, and remediation directions.

Does deep-qa work with pull requests and research reports?

Yes, deep-qa reviews pull requests, research reports, specifications, prompts, and existing code changes through independent QA dimensions. It applies adversarial critique and pre-completion verification to any software artifact requiring structured defect detection.

When should I not use automated adversarial critique for defect detection?

You should avoid automated adversarial critique when your artifact lacks defined specifications or when verification requires runtime execution without configured endpoints for cross-model review. The process relies on termination safeguards and structured inputs to effectively coordinate parallel critics.