What it does
If any judgment fails or credentials are missing, it preserves the original result order.
Agent tooling
Medical MCP asks Jev to rank fetched clinical literature against a research question.
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
If any judgment fails or credentials are missing, it preserves the original result order.
Start by writing a clear clinical research question for your literature search. The tool already includes built-in checks for every paper it finds. It automatically evaluates whether each study tests real human patients, matches your exact medical setting, and offers direct evidence rather than background reading.
A developer can hook this tool into your workflow using an access key from TypeSafe, which is a private code that lets the app communicate with the ranking service. If that connection code is missing or an automated check fails, the tool simply leaves your search results in their original order.