jkudish/jev-mcp ? reverse-engineered prompt

Reverse engineered prompt

Build me an MCP server called Jev that gives agents fast, cheap judgment tools for checking claims, screening content, finding the best match, reranking results, classifying items, comparing passages, extracting fields, and reviewing diffs.

I want it to work locally through standard MCP, and also in HTTP mode for a shared or remote setup. It should use a TypeSafe Jev API key from the environment, and require an auth token for the HTTP version unless it is only running on my machine. Please include clear setup instructions, a simple way to register it with common MCP clients, and a small skill or guide that tells agents when to use each tool instead of guessing.

Make the results come back in a typed way, with verdicts, probabilities, and confidence where that makes sense. If you need to check current docs or client setup details, look them up online.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab