Report Abuse

Basic Information

emcee is a command-line tool that exposes a Model Context Protocol (MCP) server for any web application that publishes an OpenAPI specification. It is intended for developers who want to make APIs available as tools to AI clients such as Claude Desktop and other MCP-compatible applications, enabling plugin-like access to external tools and data without writing a custom MCP server. emcee runs as a stdio JSON-RPC server, accepts an OpenAPI spec file or URL as input, and exposes selected endpoints as MCP tools. The README documents installation options (installer script, Homebrew, Docker, or build from source with Go 1.24+), basic usage and CLI flags, authentication options, and guidance for transforming OpenAPI specs before exposing them. It is positioned as a convenient bridge for services you are building that lack native MCP implementations.

Links

Categorization

App Details

Features
Generates an MCP server from an OpenAPI specification and exposes API endpoints as tools consumable by MCP clients. Implements the stdio transport using JSON-RPC 2.0 as the wire format and provides JSON-RPC methods such as tools/list and tools/call. Supports multiple authentication modes via CLI flags: bearer token, basic auth and raw Authorization header values, and optional 1Password secret references. Offers installation via an installer script, Homebrew, Docker images, or building from source with Go 1.24+. Includes CLI flags for retries, requests-per-second rate limiting, request timeout, verbosity and silent logging. Documents transforming OpenAPI specs with jq/yq or overlays to select or modify endpoints. Notes that emcee does not use auth when fetching remote OpenAPI URLs and logs to stderr for diagnostics.
Use Cases
emcee lets developers quickly expose existing web APIs as tools for AI assistants without building a bespoke MCP implementation. By pointing emcee at an OpenAPI spec you can make selected endpoints available to clients like Claude Desktop so the model can request and use external data or actions, similar to ChatGPT plugins. Authentication can be provided via CLI or 1Password references so protected APIs can be surfaced securely at runtime. The ability to transform specs before exposing them makes it easy to limit or reshape the toolset. emcee also supports standard developer workflows with Docker and build-from-source options, and debugging with the MCP Inspector to test and observe MCP interactions. It is useful for prototyping integrations and enabling model access to internal or third-party services.

Please fill the required fields*