JetBrains AI Assistant with Keeptrusts
Current JetBrains AI Assistant releases support OpenAI-compatible endpoints and custom MCP servers. Configure the model and tool paths separately.
Route model requests
- Start the gateway using Gateway Setup for IDEs.
- Open Settings → Tools → AI Assistant → Providers & API keys.
- Add an OpenAI-compatible endpoint.
- Set the endpoint to
http://127.0.0.1:41002/v1. - Enter a Keeptrusts client API token, then test the connection.
- Assign a returned model to the AI Assistant feature you intend to use.
JetBrains assigns custom models to feature groups, and some features may be unavailable if the selected model lacks the required protocol or capability. Inline completion also has model-specific requirements; do not assume that a chat model is a fill-in-the-middle completion model.
See JetBrains' custom-model documentation for the current settings and assignment rules.
Add Keeptrusts MCP tools
Open Settings → Tools → AI Assistant → Model Context Protocol (MCP), add a remote HTTP server, and configure:
- URL:
https://<published-hostname>/mcp - Header:
Authorization: Bearer <client-api-token>
Use the IDE's credential UI rather than storing a token in a project file. MCP tools supplement the selected model; they do not reroute a JetBrains-hosted model by themselves.
Verify
Run kt events tail --since 10m --follow, select the Keeptrusts model, and send
a chat request. Then inspect the MCP server status and invoke a Keeptrusts tool
as a separate check.