Jesse MCP — connect in Codex

Make sure Jesse is running and your .env is set up so the MCP server starts and prints a URL (see MCP server setup if you have not done that yet).

These steps assume you use Codex with custom MCP support. Jesse exposes a running Streamable HTTP MCP server; Codex does not need to launch Jesse with a STDIO command.

Register Jesse in Codex

1. Open Codex MCP settings and choose Connect to a custom MCP.
2. Set Name to jesse.
3. Select Streamable HTTP.
4. Use the URL Jesse printed in the terminal. It should end with /mcp (often http://localhost:9002/mcp unless you changed MCP_PORT).
5. Leave Command to launch, Arguments, Environment variables, and Environment variable passthrough empty.
6. Leave Working directory empty unless Codex requires one. If it does, use your Jesse project root.
7. Save the MCP server.

If something fails, turn on MCP_LOG_IN_TERMINAL in .env while you debug—see Environment variables for MCP.

Agent rules (optional but recommended)

Jesse automatically writes and maintains AGENTS.md at the project root every time you run jesse run, so Codex can read the rules without you pasting anything. See MCP agent rules for details on what's in the file, how the auto-sync works, and how to add your own notes without losing them on upgrade.

Test the connection

After saving the MCP server, ask Codex:

Use the Jesse MCP server and list the available tools.

If Codex cannot find the server, confirm Jesse is still running and that the URL matches the terminal output.

• OpenAI Docs MCP quickstart

Same machine, same network story

If Codex runs on Windows and Jesse runs inside WSL (or the other way around), localhost can refer to different machines. If you see connection errors, run Jesse and Codex in the same environment so they share one loopback network.