Verification & Quality Assurance

Verifies code quality with truth scoring and automatic rollback of failed changes.

11|Updated Jan 20, 2026
One-click install
npx skills add https://github.com/ishandutta2007/claude-agent-orchestration --skill verification-quality-assurance-ishandutta2007
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Verification & Quality Assurance
Source: https://github.com/ishandutta2007/claude-agent-orchestration/tree/main/.claude/skills/verification-quality
Command: npx skills add https://github.com/ishandutta2007/claude-agent-orchestration --skill verification-quality-assurance-ishandutta2007

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Agent-generated code and task outputs can silently introduce defects, security issues, or regressions. This Skill provides automated verification with quantified truth scores and instant rollback so low-quality changes never persist in your codebase. ## Core Features & Use Cases - Truth Scoring: Computes reliability metrics on a 0.0-1.0 scale for files, agents, and tasks, with trends, statistics, and exportable reports in JSON, CSV, HTML, or Markdown. - Verification Checks: Validates code correctness, security, performance, documentation, and best practices against a configurable threshold (default 0.95). - Automatic Rollback: Reverts failed changes via Git in under a second, with selective, dry-run, and backup-first modes. - Use Case: Add a verification step to your GitHub Actions pipeline that runs a JSON verification check on every pull request and fails the build when the overall truth score drops below 0.95. ## Quick Start Run a verification check on my current project and show me the truth scores for the last 7 days.

Frequently Asked Questions about Verification & Quality Assurance

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify code quality with a truth score threshold?

Run npx claude-flow@alpha verify check to score your code against correctness, security, performance, and documentation criteria. Use --threshold 0.98 to set a custom pass level, or --file to target a specific file.

How to automatically rollback code changes that fail verification?

Use npx claude-flow@alpha verify rollback --last-good to revert to the last passing state via Git in under a second. Options include --selective to keep good changes, --dry-run to preview, and --backup-first for safety.

Can I integrate truth score verification into GitHub Actions CI?

Yes, run verify check with --json output in a workflow step, then parse overallScore with jq and exit non-zero when it falls below your threshold. The verification report can be uploaded as a build artifact.

What does the truth score number actually measure?

Truth scores range from 0.0 to 1.0 and aggregate code correctness, security, performance, documentation, and best-practices checks. Scores above 0.95 are production-ready, 0.85-0.94 acceptable, and below 0.75 require immediate action.

Why is my verification check timing out on large codebases?

Increase the limit with --timeout 60s or split work using verify batch with --batch-size. A full codebase analysis typically completes in under 5 seconds, so persistent timeouts usually indicate an unusually large directory scan.