Agent♥︎Age
Catalog

io.github.Skillproofdev/skillproof

Official

by Skillproofdev · JavaScript

Check whether a Claude Code skill was tested and whether it works before installing it.

SkillProof MCP (io.github.Skillproofdev/skillproof)

This Model Context Protocol (MCP) server checks whether a Claude Code skill was tested and whether it works before installing it. It is based on SkillProof’s approach of running skill content against a real task and a no-skill baseline, returning verdicts to an agent.

🛠️ Key Features

  • Claude Code skill testing status
  • Works/non-works verdicts before installation
  • Uses SkillProof’s “no-skill baseline” comparison
  • Topics include: ai-tools, claude, claude-code, claude-skills, mcp, model-context-protocol

🚀 Use Cases

  • Ask if a tested skill exists for a specific task (e.g., markdown-to-Confluence conversion)
  • Validate outcomes such as “DIDN’T PASS” with a score below baseline
  • Surface test details from a run (e.g., “convert_markdown_to_wiki.py”)

⚡ Developer Benefits

  • Decision support for whether to install a Claude Code skill
  • Evidence-based results from tasks run by SkillProof
  • Developer-usable MCP interface for verdict retrieval

⚠️ Limitations

  • Relies on SkillProof’s testing process; skills not covered by its verdicts cannot be confirmed here

Topics

ai-toolsclaudeclaude-codeclaude-skillsmcpmodel-context-protocol
io.github.Skillproofdev/skillproof - agentage MCP Catalog