Setup
Install and configure the Jeremy MCP server for use with AI assistants.
Install
Install the MCP server globally:
npm install -g @jeremy/mcpThis provides the jeremy-mcp command.
Configure in Claude Code
Add the Jeremy MCP server to your Claude Code configuration. You can add it to either:
- Project settings --
.claude/settings.jsonin your project root - Global settings --
~/.claude/settings.json - Claude Desktop --
claude_desktop_config.json
{
"mcpServers": {
"jeremy": {
"command": "jeremy-mcp",
"env": {
"JEREMY_API_URL": "https://jeremy-app.ian-muench.workers.dev",
"JEREMY_API_KEY": "jrmy_your-api-key"
}
}
}
}Replace jrmy_your-api-key with an API key generated from the Jeremy Dashboard.
Environment Variables
| Variable | Description | Required |
|---|---|---|
JEREMY_API_URL | Base URL of your Jeremy instance | Yes |
JEREMY_API_KEY | Your API key for authentication | Yes |
If you are self-hosting Jeremy, set JEREMY_API_URL to your own instance URL.
Verify
Once configured, restart Claude Code. The assistant should now have access to the resolve-library-id and query-docs tools. You can verify by asking Claude to look up documentation for a library you have already ingested.
Other MCP Clients
The Jeremy MCP server works with any MCP-compatible client, not just Claude Code. The server communicates over stdio using the standard MCP protocol. Point your client at the jeremy-mcp command and provide the required environment variables.