What triggers the verification gate?

Any expression of satisfaction, completion, or status (such as 'Done!', 'Fixed', or 'Tests pass') triggers the requirement to provide fresh evidence.

Why is this skill necessary for AI development?

It prevents the AI from assuming work is complete based on intuition or partial data, ensuring every success claim is backed by actual, fresh command output.

What happens if a verification command fails?

The skill mandates that the actual status be reported with evidence, preventing the AI from hiding failures or moving to the next task prematurely.

Does it support Test Driven Development (TDD)?

Yes, it includes specific patterns for Red-Green verification to ensure that regression tests are valid and truly catch the intended issues.

How does it handle sub-agent reports?

It refuses to trust a sub-agent's success report blindly, requiring a manual check of the version control diff and a fresh verification run.

Rigorous Verification Guard

Name: Rigorous Verification Guard
Author: ehartye

byehartye

0•

보안 및 테스팅

Enforces evidence-based reporting by requiring fresh verification evidence before any claims of task completion or success.

This skill acts as a critical gatekeeper for high-integrity development, preventing Claude from making premature or unverified claims about code status. By mandating the 'Iron Law' of fresh evidence, it ensures that every assertion of success—whether it's a passing test, a clean lint, or a successful build—is backed by real-time command output. It eliminates common pitfalls like 'confidence-based' reporting or trusting outdated results, fostering total transparency and technical honesty throughout the software development lifecycle.

주요 기능

01Enforces a strict 'Evidence Before Claims' policy for all status updates

02Detects and blocks rationalization language like 'should pass' or 'seems to'

03Implements a mandatory gate function to identify and run proof commands

04Validates sub-agent output through independent verification and diff analysis

050 GitHub stars

06Supports TDD workflows with required Red-Green cycle verification

사용 사례

01Auditing complex task completion to ensure all line-by-line requirements are met

02Verifying bug fixes by reproducing symptoms and confirming resolution with logs

03Ensuring all tests and build steps pass before committing code or creating pull requests

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ehartye/hartye-superpowers verification-before-completion

For use in Claude.ai and ChatGPT

Download Skill