What is the primary output of the WF Verifier skill?

It generates a standardized verification report located at design-docs/plans/{name}/{name}_verification.md that categorizes findings into critical, warning, and suggestion levels.

What is 'LARP detection' in the context of this skill?

LARP (Live Action Role Play) detection is a specialized check that hunts for fake, performative, or redundant code that looks functional but adds no real value.

Does this skill fix code issues automatically?

Yes, one of its specialized dimensions (Code Cleanliness) is designed to find and fix style issues and AI slop as it performs the review.

How is the WF Verifier launched?

It is launched via the Claude Code interface using the command: claude --agent wf:verifier.

How does WF Verifier handle documentation debt?

It treats stale or missing documentation as a high-severity bug, checking ARCHITECTURE.md, living docs, and plans against the actual code built.

WF Verifier

Name: WF Verifier
Author: JSai23

byJSai23

•

Security & Testing

Conducts comprehensive end-of-cycle quality reviews, architectural audits, and documentation health checks to ensure production readiness.

WF Verifier acts as a final gatekeeper in the development lifecycle, specialized in assessing code integrity, design hindsight, and documentation accuracy. By delegating specific checks to sub-agents for style, design, 'LARP' detection (identifying performative or fake code), and production quality, it synthesizes a detailed verification report. It ensures that system documentation like ARCHITECTURE.md and living docs remain synced with the actual implementation, preventing both technical and documentation debt before code reaches production.

Key Features

01Automated documentation health auditing for living docs and ARCHITECTURE.md

02Specialized LARP detection to identify performative code or AI-generated slop

03Delegated sub-agent execution for specialized architectural and quality checks

04Detailed verification report generation with specific file and line references

051 GitHub stars

06Multi-dimensional verification covering design, style, quality, and integrity

Use Cases

01Synchronizing system architecture maps with actual implementation to prevent doc drift

02Performing deep-dive architectural reviews of newly completed modules in hindsight

03Finalizing a feature branch with a comprehensive quality and documentation audit before merge

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jsai23/claude-tooling wf-verifier

For use in Claude.ai and ChatGPT

Download Skill