code-verification

Identify failure modes in code changes through adversarial verification.

3|1|Updated Apr 1, 2026

Search Tags:#testing #verification #robustness #quality-assurance #adversarial #code-changes #breakage

Ready to run

npx skills add https://github.com/bhaumikmaan/claude-code-master-skills --skill code-verification-bhaumikmaan

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: code-verification
Source: https://github.com/bhaumikmaan/claude-code-master-skills/tree/main/skills/code-verification
Command: npx skills add https://github.com/bhaumikmaan/claude-code-master-skills --skill code-verification-bhaumikmaan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Identifies and exposes failure modes in code changes through adversarial verification.

Core Features & Use Cases

Adversarial verification strategies tailored to various change types (frontend, backend, CLI, infrastructure, library).
Structured PASS/FAIL/PARTIAL verdicts with evidence to guide improvements.
Comprehensive verification workflow including readme, tests, and checks before task completion.

Quick Start

Run the verification workflow on your latest commit to surface edge cases and verify robustness.

Frequently Asked Questions about code-verification

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

FAQPage Schema

How do I verify code changes for edge cases and breakage?▼

Verify code changes for edge cases by running an adversarial verification workflow that probes frontend, backend, CLI, infrastructure, and library updates to expose failure modes and performance regressions.

What is adversarial verification in software testing?▼

Adversarial verification is a testing method that rigorously probes code changes to identify failure modes, targeting edge cases and integration points to ensure robustness before delivering structured PASS, FAIL, or PARTIAL verdicts.

How do I perform adversarial code probing on infrastructure changes?▼

Perform adversarial code probing on infrastructure changes by applying tailored verification strategies that test integration points and edge cases, collecting evidence to guide improvements and output structured PASS, FAIL, or PARTIAL verdicts.

Does adversarial verification work for frontend, backend, and CLI changes?▼

Yes, adversarial verification works for frontend, backend, CLI, infrastructure, and library changes, applying tailored strategies to surface performance regressions and integration failures across various change types.

What's the best way to surface failure modes in new code commits?▼

The best way to surface failure modes in new commits is running a comprehensive verification workflow that includes readme, tests, and checks, applying adversarial probing to deliver evidence-backed PASS, FAIL, or PARTIAL verdicts.