What is the target test coverage for the verification loop?

The skill defaults to a target of 80% code coverage, reporting the total tests passed and failed alongside the coverage percentage.

How does the security scan phase work?

The skill performs targeted scans using pattern matching to identify potential hardcoded secrets (like API keys) and leftover console logs or debug statements in your source code.

What languages does the Verification Loop support?

It currently includes pre-configured commands for TypeScript, JavaScript, and Python projects, but the verification framework can be adapted for any programming language.

When is the best time to invoke this skill?

You should invoke it after completing a feature, before creating a PR, or every 15-20 minutes during long sessions to maintain a high quality bar.

Does the skill automatically fix the errors it finds?

It focuses on identification and reporting. Once the Verification Report highlights issues, Claude or the developer can systematically address each one before re-running the loop.

Verification Loop

Name: Verification Loop
Author: affaan-m

byaffaan-m

•

112,919

•

安全与测试

Standardizes and automates a multi-phase quality assurance process for code changes within Claude Code sessions.

The Verification Loop skill provides a robust framework for validating code integrity through six distinct phases: build verification, type checking, linting, comprehensive testing with coverage analysis, security scanning for secrets, and a final git diff review. By enforcing strict quality gates and generating a structured verification report, it ensures that every code change is production-ready and free of regressions before reaching the pull request stage, making it an essential tool for maintaining high code standards in AI-assisted development.

主要功能

01112,919 GitHub stars

02Six-phase quality assurance pipeline covering builds, types, and tests

03Built-in security scanning for leaked API keys and sensitive debugging logs

04Standardized verification reporting with clear PASS/FAIL indicators

05Cross-language support for TypeScript, JavaScript, and Python environments

06Automated test suite execution with coverage threshold monitoring

使用场景

01Finalizing code changes and ensuring readiness before opening a pull request

02Validating complex refactors to ensure no regressions or type errors were introduced

03Implementing periodic quality checks during long-running AI development sessions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add affaan-m/everything-claude-code verification-loop

For use in Claude.ai and ChatGPT

Download Skill