analyzer
01Agent Codex
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
10 2d ago A 0 tokens
copy · 100% MIT
5 tagged scenario, measured the same way as everything else here.
Agent Codex
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
Agent Claude Code
Reviews supporting files (scripts, references, assets) that sit next to a SKILL.md. Use proactively after adding or changing any file in skills/ / other than SKILL.md itself.
Agent Claude Code
Clean-room tester that runs one realistic task using only an installed skill as documentation, then reports what it did. Invoked by the /skills:validate command; do not delegate ordinary work to it.