Does this work with TDD workflows?

Yes, it specifically encourages Red-Green cycles to confirm that tests fail when they should and pass only after the fix is applied.

How does it handle AI rationalization?

It identifies 'red flag' phrases like 'seems to work' or 'should be fixed' and forces the agent to stop and run a verification command before proceeding.

Can I use this for PR preparation?

Absolutely. It is designed to be the final gate before committing code, creating a Pull Request, or moving to the next task in a workflow.

What kind of commands does this skill require?

It requires running relevant tests, linters, build scripts, or specific reproduction scripts for bugs to provide empirical evidence of success.

Why is this skill important for AI coding?

It prevents the AI from making assumptions about code success, ensuring every claim of completion is backed by real, fresh command output rather than optimistic predictions.

Rigorous Verification Guard

Name: Rigorous Verification Guard
Author: zhengxuyu

byzhengxuyu

0•

보안 및 테스팅

Enforces a strict evidence-based verification protocol before any code changes are claimed as complete or successful.

This skill implements a mandatory gate function for Claude Code, ensuring that no task is marked as finished without fresh, empirical evidence. By requiring the execution and analysis of actual verification commands—such as tests, builds, or linters—it eliminates guesswork and prevents false positives. It is particularly useful for maintaining high code quality, preventing regressions, and building trust between the AI agent and the human developer by prioritizing actual command output over optimistic assertions like 'should work' or 'seems correct.'

주요 기능

01Mandatory 'Gate Function' requiring command execution before any success claim

02Support for Red-Green TDD verification cycles to ensure regression test validity

03Comprehensive checklists for bug fixes, builds, linting, and requirement mapping

04Independent verification of agent delegation via VCS diffs and build logs

05Detection and prevention of rationalization language like 'should' or 'probably'

060 GitHub stars

사용 사례

01Ensuring all unit and integration tests pass before opening a Pull Request

02Validating that a project remains linter-clean and buildable after a complex refactor

03Verifying a bug fix via the Red-Green cycle to confirm the fix actually addresses the symptom

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zhengxuyu/claude-superpowers verification-before-completion

For use in Claude.ai and ChatGPT

주요 기능

01Mandatory 'Gate Function' requiring command execution before any success claim

02Support for Red-Green TDD verification cycles to ensure regression test validity

03Comprehensive checklists for bug fixes, builds, linting, and requirement mapping

04Independent verification of agent delegation via VCS diffs and build logs

05Detection and prevention of rationalization language like 'should' or 'probably'

060 GitHub stars

사용 사례

01Ensuring all unit and integration tests pass before opening a Pull Request

02Validating that a project remains linter-clean and buildable after a complex refactor

03Verifying a bug fix via the Red-Green cycle to confirm the fix actually addresses the symptom

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zhengxuyu/claude-superpowers verification-before-completion

For use in Claude.ai and ChatGPT