How does the automatic rollback feature work?

If a verification check falls below the configured threshold (defaulting to 0.95), the skill uses Git integration to instantly revert the changes to the last known good state.

Can I integrate this with GitHub Actions?

Yes, the skill provides structured JSON output and CLI commands specifically designed for integration into CI/CD pipelines to act as a quality gate.

What is a Truth Score in this skill?

A Truth Score is a real-time reliability metric on a 0.0 to 1.0 scale that evaluates code, agents, and tasks based on correctness, security, and logic flow.

What specific criteria does the verification system evaluate?

It evaluates five key areas: Code Correctness (syntax/logic), Best Practices (style/SOLID), Security (vulnerabilities/secrets), Performance (complexity/memory), and Documentation quality.

Is the accuracy threshold customizable?

Absolutely. You can set custom thresholds via the CLI using the --threshold flag or define environment-specific levels (e.g., 0.99 for production) in your config file.

Verification & Quality Assurance

Name: Verification & Quality Assurance
Author: while-basic

bywhile-basic

0•

Security & Testing

Ensures codebase reliability and agent output quality through automated truth scoring, multi-dimensional verification checks, and instant rollbacks.

The Verification & Quality Assurance skill provides an advanced reliability layer for Claude Flow, utilizing a proprietary truth-scoring system to evaluate the accuracy and quality of AI-generated code. It automates rigorous checks for syntax correctness, security vulnerabilities, performance bottlenecks, and adherence to best practices, featuring a built-in safety mechanism that triggers automatic rollbacks when quality thresholds are not met. This tool is essential for developers integrating AI agents into production pipelines who require statistical confidence, real-time monitoring, and verifiable output standards.

Key Features

01Interactive dashboard with trend analysis, confidence intervals, and live updates

02Automated verification checks for security, performance, and best practices

03Deep CI/CD integration with exportable JSON, HTML, and CSV reporting

040 GitHub stars

05Real-time Truth Scoring (0.0-1.0 scale) for code and agent task reliability

06Smart Rollback system that reverts changes failing the 0.95 accuracy threshold

Use Cases

01Validating AI-generated code for production readiness before merging pull requests

02Automating quality gates in CI/CD pipelines with customizable truth thresholds

03Monitoring long-term agent performance and codebase health via statistical dashboards

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add while-basic/c-flow verification-quality

For use in Claude.ai and ChatGPT

Download Skill