Skill Judge
Skill Judge analyzes text for factual accuracy and bias, helping users verify information and assess arguments critically.
Install on your platform
We auto-selected Claude Code based on this skill’s supported platforms.
Run in terminal (recommended)
claude mcp add gohypergiant-skill-judge npx -- -y @trustedskills/gohypergiant-skill-judge
Or manually add to ~/.claude/settings.json
{
"mcpServers": {
"gohypergiant-skill-judge": {
"command": "npx",
"args": [
"-y",
"@trustedskills/gohypergiant-skill-judge"
]
}
}
}Requires Claude Code (claude CLI). Run claude --version to verify your install.
About This Skill
What it does
Skill Judge analyzes agent skills against official specifications and patterns derived from over seventeen examples. It's designed to evaluate how well an agent skill is performing, acting as a knowledge externalization mechanism rather than a tutorial or training process. The core purpose of this skill is to identify the "knowledge delta" - what new expertise it provides beyond what the underlying AI model already knows.
When to use it
- When assessing the quality and effectiveness of an agent skill's implementation.
- To ensure agent skills adhere to established specifications and patterns.
- For evaluating newly created or modified agent skills.
- During development cycles to identify areas for improvement in agent skill design.
Key capabilities
- Evaluates agent skills against official specifications.
- Identifies knowledge gaps ("knowledge delta") compared to the AI model's existing knowledge.
- Focuses on expert-only knowledge, decision trees, trade-offs, and domain-specific thinking frameworks.
Example prompts
- "Evaluate this skill for adherence to the MCP specification."
- "Assess this agent skill’s performance against established patterns."
- “What is the knowledge delta of this Skill?”
Tips & gotchas
- A Skill Judge's value lies in providing expert-only knowledge that the AI model doesn't already possess. Avoid explaining basic concepts or common practices, as this wastes valuable context window space.
- This skill is not intended to provide tutorials or training materials; it’s for evaluating existing skills.
Tags
TrustedSkills Verification
Unlike other registries that point to live repositories, TrustedSkills pins every skill to a verified commit hash. This protects you from malicious updates — what you install today is exactly what was reviewed and verified.
Security Audits
| Gen Agent Trust Hub | Pass |
| Socket | Pass |
| Snyk | Pass |
🌐 Community
Passed automated security scans.