A bridge between Streamable HTTP and stdio MCP transports, enabling flexible MCP server connectivity.
mcp-proxy is easy to set up with strong trust signals. Check agent compatibility and use-case fit before adding it to your workflow.
gh repo view sparfenyuk/mcp-proxy --webOpen the official repository or website.
Check the README for package manager, auth, and platform requirements.
Try it in a small test task inside your agent workflow.
mcp-proxy lets you connect MCP clients (like Claude Desktop) to remote servers using different transport protocols. It can convert stdio to SSE/StreamableHTTP or vice versa, making it easy to use remote MCP servers with local tools.
mcp-proxy is a tool that bridges different MCP (Model Context Protocol) transports. It supports two modes: 1) stdio to SSE/StreamableHTTP: allows local MCP clients (e.g., Claude Desktop) to connect to remote MCP servers over SSE or StreamableHTTP. 2) SSE to stdio: allows remote SSE clients to connect to local stdio-based MCP servers. The proxy handles authentication via headers or OAuth2, supports custom headers, and can be configured via command-line arguments or a config file. It is available as a Python package (PyPI), Docker image, or from source. The project is actively maintained with 2576 stars and 242 forks.
Strong trust signals; still review the README and permissions before production use.
Last commit was about 4 days ago.
2576 GitHub stars indicate community interest.
48 open issues signal maintenance load.
MIT license detected.
Connect Claude Desktop to a remote MCP server over SSE
Expose a local stdio MCP server to remote clients via SSE
Bridge StreamableHTTP servers to stdio-only MCP clients
Add authentication (OAuth2, custom headers) to MCP connections
Containerize MCP proxy for deployment in Docker environments
Exposing a proxy to the network may introduce security risks if not properly configured
OAuth2 credentials and custom headers are passed in plain text; use HTTPS in production
The proxy does not enforce encryption; ensure secure transport (TLS) when used over networks
2,576
Stars
242
Forks
48
Issues
MIT
License
A terminal dashboard to monitor AI coding agents like Claude Code and Codex CLI in real-time.
AI agent security scanner that detects vulnerabilities in agent configurations, MCP servers, and tool permissions.
A full-stack AI Red Teaming platform for securing AI ecosystems with comprehensive scanning and evaluation.
3 security/trust notes recorded.
Setup difficulty is 2/5.