mcp-proxy is a self-hosted tool for connecting AI clients and MCP servers that use different connection types. It's for developers who need a stdio client, such as Claude Desktop, to reach an SSE or Streamable HTTP server, or who want remote clients to access a local stdio server. The proxy runs on your own machine or server.
It works in both directions. A desktop client can communicate through the proxy with a remote MCP service, while a local MCP server can become accessible over HTTP without changing its own stdio interface. Requests to a remote service go to that service; running the proxy locally doesn't keep those requests on your machine.
One proxy can expose multiple named MCP servers, each at a separate URL. That makes it useful for a self-hosted setup with several tools, rather than a separate proxy for every server. It also supports bearer tokens and OAuth2 authentication when connecting to remote services, plus controls for which browser origins can access its server endpoints. Streamable HTTP connections can use stateless mode.
The project is open source under the MIT license and written in Python. It runs on macOS, Windows and Linux, with Docker images available for Linux on AMD64 and ARM64. Compatible server examples include mcp-server-fetch and @modelcontextprotocol/server-github.
Claim this page and we'll verify you by hand. mcp-proxy gets the verified badge, and you can upgrade the listing to be featured on localhosted. Proud to be listed? Put our badge on your site.
Want more people to find mcp-proxy?Promote it
Something wrong or outdated on this page?
42Updated 12 hours ago
macOS · Windows · Linux · Docker#Agent Skills#Code execution#Distributed execution
Code Buddy is an AI coding assistant for developers who want a terminal agent running on their own machine, with a choice of local or cloud models. It reads repositories, edits code and runs commands on Linux, macOS and Windows. Ollama keeps model inference local without an API key or account; cloud providers send model requests off the machine. Routing includes automatic provider failover.
27.3KUpdated 20 hours agoMIT
macOS · Windows · Linux#Code execution#MCP#Tool calling
1.1KUpdated 7 hours agoMIT
macOS · Windows · Linux#Agent Skills#MCP#Persistent memory
68.2KUpdated 1 day agoMIT
macOS · Windows · Linux#MCP#Works offline
4.4KUpdated 19 hours agoMIT
macOS · Windows · Linux · Web · JetBrains#Agent Client Protocol#Code execution#Git integration
155.4KUpdated 1 day agoMIT
macOS · Windows · Docker · Web#LLM tracing#MCP#Multi-agent workflows
Cua gives AI agents access to computers they can inspect and operate, with tools for desktop automation, local virtual machines, and hosted fleets. It's for developers building agents that work across native apps and browsers, or evaluating how well those agents complete computer tasks. You bring the agent and model.
deja-vu gives coding assistants a shared memory of work already recorded on your machine. It searches sessions from before you installed it, so developers can recover an old fix or carry context between Claude Code, Codex CLI, Cursor and opencode without starting a separate collection of notes.
Docling is an MIT-licensed, open source document parser for developers turning files into structured content for search and AI applications. It runs locally on macOS, Linux, and Windows, including in air-gapped environments. Its PDF processing identifies page layout and reading order, extracts tables, code, and formulas, and classifies images.
gptme is a self-hosted AI agent that works directly in your terminal, with access to your files and installed tools. It's for developers who want a coding assistant in their own environment, and people who want an agent for data analysis or other knowledge work. The software is free under the MIT license and doesn't require a gptme account.
Langflow is a visual builder for developers creating AI agents and retrieval-augmented generation (RAG) applications. You can run it locally or on your own server, with Docker support and desktop apps for Windows and macOS. The open-source software uses the MIT license. A hosted cloud offering provides a separate deployment option.