When is the best time to invoke this skill?

It is best used after completing a feature, before creating a PR, or periodically every 15-20 minutes during long development sessions to prevent error accumulation.

Can I customize the test coverage threshold?

Yes, the skill is designed to target an 80% minimum coverage by default, but this can be adjusted within your specific project configuration.

What programming languages does the Verification Loop support?

The skill includes built-in patterns for TypeScript, JavaScript, and Python, but its logic can be easily adapted for C++, Rust, or any language with command-line build and test tools.

How does the security scan phase work?

It performs targeted grep searches for sensitive patterns like API keys (e.g., 'sk-') and checks for common production-ready issues like leftover console.log statements.

Does this skill replace a project's CI/CD pipeline?

No, it serves as a 'pre-flight' check to catch errors locally within the Claude Code session, reducing CI failure cycles and improving code review speed.

Verification Loop

Name: Verification Loop
Author: ysyecust

byysyecust

0•

セキュリティとテスト

Automates a comprehensive multi-phase quality gate system to verify code integrity through builds, tests, and security scans.

The Verification Loop skill provides a rigorous framework for Claude Code to validate development work before it reaches production or code review. By executing a sequence of critical quality gates—including build validation, static type checking, linting, test suite execution with coverage metrics, and security audits—it ensures that every code change is robust, secure, and project-compliant. This skill is essential for maintaining high standards during feature completion, large-scale refactoring, or as a routine 'pre-flight' check before submitting pull requests.

主な機能

01Proactive security scanning for leaked secrets and accidental debug logs

020 GitHub stars

03Multi-phase validation covering build status, types, and linting standards

04Integrated git diff analysis to identify unintended side effects

05Standardized verification reporting for clear 'Ready for PR' status updates

06Automated test suite execution with detailed coverage threshold reporting

ユースケース

01Validating the stability of complex refactors across large codebases

02Ensuring code quality and passing all tests before opening a Pull Request

03Setting recurring mental checkpoints during long AI-assisted coding sessions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ysyecust/everything-claude-code verification-loop

For use in Claude.ai and ChatGPT

Download Skill