How does the automatic rollback feature function?

The skill utilizes Git integration to instantly revert the codebase to the last known 'good' state if a verification check falls below your specified threshold (default is 0.95).

Can I integrate these checks into GitHub Actions?

Yes, the skill provides structured JSON output and CLI commands designed specifically for integration with GitHub Actions, GitLab CI, and other automation workflows.

What specific quality metrics does the skill analyze?

It evaluates code correctness (syntax/logic), security (vulnerabilities/secrets), performance (complexity), best practices (SOLID/style), and documentation completeness.

What is a Truth Score in this skill?

A Truth Score is a real-time reliability metric on a scale of 0.0 to 1.0 that evaluates the correctness, security, and performance of code changes or agent tasks.

Is there a visual way to track quality over time?

Yes, the skill includes an interactive web-based dashboard that provides real-time updates, trend visualizations, and historical performance charts for your project.

Verification & Quality Assurance

Name: Verification & Quality Assurance
Author: whrit

bywhrit

0•

セキュリティとテスト

Ensures codebase reliability and agent output accuracy through real-time truth scoring, automated quality checks, and intelligent Git-based rollbacks.

The Verification & Quality Assurance skill provides a rigorous framework for maintaining high-integrity codebases when working with AI agents. It implements a sophisticated 'Truth Scoring' system that evaluates code on a 0.0 to 1.0 scale across dimensions like security, performance, and best practices. By setting strict accuracy thresholds and enabling automatic rollbacks for failing changes, this skill bridges the gap between autonomous agent output and production-grade reliability, offering developers real-time monitoring dashboards and seamless CI/CD integration.

主な機能

01Intelligent Git-based rollback system to instantly revert low-quality changes.

02Real-time Truth Scoring (0.0-1.0) to quantify code and agent task reliability.

03Multi-format reporting (JSON, HTML, CSV) for seamless CI/CD pipeline integration.

04Interactive web dashboards for live monitoring of codebase health and quality trends.

05Automated verification checks covering security, performance, and SOLID principles.

060 GitHub stars

ユースケース

01Automating quality gates in CI/CD pipelines to block merges that fail verification thresholds.

02Monitoring and auditing AI agent performance during large-scale refactoring or code generation.

03Ensuring security compliance and best practice adherence in collaborative development environments.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add whrit/agent-flow verification-quality

For use in Claude.ai and ChatGPT

Download Skill