analyzer
01Agent Claude Code
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
5 1mo ago A 0 tokens
copy · 100% Apache-2.0
Run AI agents fully autonomously on a filesystem directory — MCP servers and skills enabled, with zero human approvals or permission prompts — safely isolated in a hardened Docker container (or locally).
Agent Claude Code
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
Agent Claude Code
Compare two outputs WITHOUT knowing which skill produced them.
Agent Claude Code
Evaluate expectations against an execution transcript and outputs.