MCP HTTP-stdio Proxy
A Node.js proxy that enables communication between stdio-based clients and HTTP-based Model Context Protocol servers. It facilitates protocol translation, session management, and authentication to allow remote HTTP servers to function as local stdio servers.
README
MCP HTTP-stdio Proxy
A Node.js implementation of a Model Context Protocol (MCP) proxy that enables communication between stdio-based clients (like Antigravity) and HTTP-based MCP servers (running with streamableHttp).
Overview
This proxy allows you to use remote or HTTP-hosted MCP servers as if they were local stdio servers. It handles the JSON-RPC communication, session management, and optional authentication seamlessly.
graph LR
Client[Antigravity / Stdio Client] -->|stdio| Proxy[MCP Proxy]
Proxy -->|HTTP JSON-RPC| Upstream[Upstream MCP Server]
Features
- Protocol Translation: Bridges MCP over stdio to MCP over HTTP.
- Session Management:
- Automatically handles upstream initialization.
- Persists session IDs in
~/.mcp-session-cachefor continuity. - Automatic retries on session expiration or "Not Initialized" errors.
- Dynamic Tool List: Automatically injects an
updatetool to manually refresh the upstream connection and clear the session cache. - Broad Support: Proxies tools, resources, prompts, completions, and logging levels.
- Authentication: Supports optional authorization headers for secure upstream servers.
- Timeout Support: Configurable request timeouts.
Installation
- Clone or copy the project files.
- Install dependencies:
npm install
Configuration
The proxy can be configured using environment variables or a .env file:
| Variable | Description | Default |
|---|---|---|
UPSTREAM_MCP_URL |
The URL of the target HTTP MCP server. | http://localhost:8080/mcp |
UPSTREAM_AUTH |
Optional authentication header value (e.g., Bearer ...). |
(none) |
MCP_TIMEOUT_MS |
Timeout for upstream requests in milliseconds. | 15000 |
MCP_SESSION_ID |
Manually specify an initial session ID. | (none) |
Usage
Direct Execution
You can run the proxy directly and optionally pass the upstream URL as the first argument:
node mcp-proxy.js https://your-remote-mcp-server.com/mcp
Run using npx (from GitHub)
You can run the proxy directly without cloning using npx:
npx -y kiennh/mcp-http-stdio https://your-remote-mcp-server.com/mcp
Run using Docker
The proxy is available as a Docker image:
docker run -i --rm \
-e UPSTREAM_MCP_URL="https://your-remote-mcp-server.com/mcp" \
-e UPSTREAM_AUTH="Bearer YOUR_TOKEN" \
ghcr.io/kiennh/mcp-http-stdio:latest
Integration with Antigravity
To use this proxy with Antigravity, add it to your mcp_config.json:
{
"mcpServers": {
"my-remote-server": {
"command": "npx",
"args": ["-y", "kiennh/mcp-http-stdio", "https://api.example.com/mcp"],
"env": {
"UPSTREAM_AUTH": "Bearer YOUR_TOKEN"
}
}
}
}
Custom 'update' Tool
The proxy automatically adds a tool named update to the list of available tools. Calling update will:
- Clear the local session cache (
~/.mcp-session-cache). - Force a re-initialization of the upstream server on the next request.
- This is useful if the upstream server's tools have changed or the session has become desynchronized.
License
ISC
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.