The Jev MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Jev account. It provides structured and secure access to available Jev models and state evaluation, so your agent can discover models, evaluate shared state, ask named Noul, Choice, or Score questions, and assess multiple questions in parallel on your behalf.
- Model discovery: Have your agent list the Jev models and aliases available to your account before choosing one for an evaluation.
- Structured state evaluation: Direct your agent to evaluate a shared state using clearly named questions that fit your decision task.
- Flexible question formats: Instruct your agent to use Noul, Choice, or Score questions based on the kind of assessment you need.
- Parallel assessments: Let the agent evaluate several named questions against the same state in parallel to gather related results in one request.