About codemcp
codemcp is an MCP command-line tool and proxy server for developers who need to inspect and invoke tools across MCP servers. It supports search, description, and execution, configurable passthrough patterns, and TypeScript stubs for models.
Key features
- 01Import server configuration from MCP clients
- 02List tools and their pin status
- 03Generate TypeScript stubs for model context
- 04Proxy search, description, and tool execution
- 05Support configured passthrough patterns
Use cases
- 01Collect MCP server configuration from clients
- 02Inspect tools exposed by connected servers
- 03Review TypeScript interfaces available to a model
- 04Invoke tools across servers through a proxy
MCP support
codemcp can run as a stdio proxy server, advertising search, describe, and execute to MCP clients, along with configured passthrough patterns.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
4Saved tool list · updated 5d ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| limit | integer | No | |
| query | string | Yes | Keywords, a tool name fragment, or a short task description. |
Details
From the project READMEcodemcp
The codemcp command and the MCP server behind it.
codemcp import --write # discover servers from your MCP clients into codemcp.config.json
codemcp tools # connect and list every tool with its pin status
codemcp stubs --detail full # print the TypeScript the model will see
codemcp serve # run the proxy over stdio
Point your MCP client at npx codemcp-cli serve. It advertises three tools, search, describe and execute, plus any passthrough patterns from the config.
Programmatic use:
import { Proxy } from "codemcp-cli";
const proxy = new Proxy({ config: { servers: { shop: { command: "node", args: ["catalog-server.mjs"] } } } });
await proxy.start();
await proxy.server.connect(transport);
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.