Verified record · from the answer index
CompletionKit
mcpofficialtested · static-verifiedscore 52/100signed passport ✓published 103d ago · verified 2026-10-243 GitHub stars
Prompt evals over MCP: run a prompt on your dataset, score each output 1-5 with an LLM judge.
registry:com.completionkit/evals · published by homemade-software-inc · source repository