---
id: mcpo
version: "0.0.20"
license: unclear
license_treatment: unclear
maintenance: active
---

# mcpo — A simple, secure MCP-to-OpenAPI proxy server

License: unclear · Maintenance: active · Downloads: 290.0K/mo

## What it is and what it does

mcpo is a proxy server that translates MCP (Model Context Protocol) servers—which typically communicate over stdio—into standard HTTP/OpenAPI endpoints. This solves a fundamental compatibility gap: MCP servers are inherently insecure and incompatible with tools expecting REST APIs, but mcpo bridges that gap by wrapping any MCP server command and exposing its tools via FastAPI with auto-generated interactive documentation, authentication, and standard HTTP semantics.

You run mcpo with an MCP server command (or point it to an SSE or HTTP-based MCP endpoint), and it immediately serves an OpenAPI schema at `/docs` and proxy routes for each tool. It supports single servers via command-line flags, multiple servers via a Claude Desktop–compatible config file with hot-reload, and OAuth 2.1 for protected endpoints. The proxy adds security (API keys, auth), stability (HTTP error handling), and interoperability (any OpenAPI client can now call your MCP tools) without requiring custom glue code.

Use it for:

- Integrate MCP tools (like mcp-server-time) with LLM agents or applications that only understand OpenAPI, enabling tool use without protocol translation.
- Expose a suite of MCP servers behind a single HTTP gateway with per-tool routing and authentication, suitable for multi-tenant or shared environments.
- Serve MCP tools behind a reverse proxy or subpath in production using `--root-path`, with hot-reload config for zero-downtime updates.
- Provide interactive API documentation and testing UI for MCP tools via the auto-generated `/docs` endpoint, improving discoverability and debugging.
- Secure stdio-based MCP servers by wrapping them in an authenticated HTTP layer with API keys or OAuth 2.1, preventing unauthorized access.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Exposes MCP (Model Context Protocol) servers as OpenAPI-compatible HTTP endpoints, making them accessible to standard REST clients and LLM agents without custom protocol handling.

Yes. mcpo solves a real interoperability problem (MCP servers are stdio-only, most tools expect OpenAPI) with low install friction, active maintenance, and no known vulnerabilities. The license treatment is unclear, so verify the actual license before use. For any workflow where you need to expose MCP tools to REST clients or LLM agents, this is the right tool.

## Install

pip install mcpo
uv add mcpo
poetry add mcpo

## Installing mcpo

Before you install: Low friction: pure Python wheel with no compiled dependencies. Active maintenance (latest release 2026-02-27). Depends on well-established libraries (FastAPI, Uvicorn, Pydantic, Click, Typer) with no known security issues.

License in practice: License treatment is unclear—no SPDX identifier or raw license text provided in metadata. Verify the actual license (description mentions MIT) before use in proprietary or restricted contexts.

Quickstart:

pip install mcpo
mcpo --port 8000 --api-key "secret" -- your_mcp_server_command
# Access OpenAPI docs at http://localhost:8000/docs

Requires Python 3.11 or later.

Verify before relying:

- License metadata shows 'unclear' treatment despite description claiming MIT—confirm actual license terms before deployment.
- No documented security audit or threat model for the proxy layer itself; verify authentication and token handling for production use.
- OAuth 2.1 support mentioned in description but not reflected in runtime dependencies—confirm implementation completeness.

## Package facts

- License: not declared (unclear)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 290.0K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

mcp to openapi proxy, model context protocol http server, mcp rest api gateway, expose mcp tools as http, mcp openapi bridge, llm agent tool integration, mcp server http wrapper, mcp-proxy, openapi-gateway, llm-integration

[View on SkillFed](https://skillfed.io/packages/mcpo) · [View on PyPI](https://pypi.org/project/mcpo/)