Does it provide visual reports?

Yes, it includes an interactive web-based dashboard that visualizes quality trends, agent performance, and task history with real-time WebSocket updates.

Can I customize the accuracy threshold?

Absolutely. You can set global, environment-specific (development vs. production), or even file-specific thresholds to match your team's strictness requirements.

Can I use this skill in my CI/CD pipeline?

Yes, the skill provides JSON and CSV export capabilities, allowing you to integrate quality checks directly into GitHub Actions, GitLab CI, or other automation workflows to block low-quality commits.

What is a Truth Score in this skill?

A Truth Score is a reliability metric ranging from 0.0 to 1.0 that evaluates the correctness of code changes based on syntax validation, logic flow, security, and adherence to best practices.

How does the automatic rollback feature work?

If a verification check fails or the truth score falls below the configured threshold (default 0.95), the skill triggers a Git-based rollback to instantly revert the codebase to its last known good state.

Verification & Quality Assurance

Name: Verification & Quality Assurance
Author: zellycloud

byzellycloud

0•

Security & Testing

Ensures code reliability and agent output quality using real-time truth scoring and automated rollback systems.

The Verification & Quality Assurance skill is a specialized framework for Claude Flow designed to maintain high-integrity codebases through rigorous automated validation. It implements a unique truth-scoring system that measures the reliability of AI agent outputs on a 0.0-1.0 scale, enforcing a strict 0.95 accuracy threshold by default. By combining deep code analysis—covering security, performance, and best practices—with instant Git-based rollbacks for failed checks, this skill provides a safety net for developers using AI agents in production environments, ensuring that every generated change is verified, documented, and reliable.

Key Features

01Automated verification checks for security, syntax, and performance logic

02Real-time Truth Scoring (0.0-1.0) for measuring agent and task reliability

030 GitHub stars

04Native CI/CD integration with support for GitHub Actions and GitLab CI

05Interactive web dashboard for live monitoring of codebase health and trends

06Instant Git-based rollback for any changes falling below quality thresholds

Use Cases

01Monitoring and benchmarking the performance of different AI coding agents

02Enforcing quality gates for AI-generated code in production repositories

03Automating regression testing and security audits during large-scale refactors

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zellycloud/zyflow verification-quality

For use in Claude.ai and ChatGPT

Download Skill