MCP BridgeKit

v1.0.0Developer Toolsstable

Embeddable MCP stdio → HTTP bridge with background jobs & live dashboard. Survives Vercel/Cloudflare 30s timeouts. Now scales to 100+ users.

agentic-aibridgekitfastapimcpmcp-gateway
Share:
60
Stars
0
Downloads
0
Weekly
0/5

What is MCP BridgeKit?

MCP BridgeKit is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to embeddable mcp stdio → http bridge with background jobs & live dashboard. survives vercel/cloudflare 30s timeouts. now scales to 100+ users.

Embeddable MCP stdio → HTTP bridge with background jobs & live dashboard. Survives Vercel/Cloudflare 30s timeouts. Now scales to 100+ users.

This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Embeddable MCP stdio → HTTP bridge with background jobs & li

Use Cases

Embeddable stdio to HTTP bridge for MCP
Serverless compatibility with background jobs
mkbhardwas12

Maintainer

LicenseMIT
Languagepython
Versionv1.0.0
UpdatedMay 7, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx mcp-bridgekit

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use MCP BridgeKit

MCP BridgeKit is a Python FastAPI application that acts as an HTTP bridge between web applications and MCP stdio servers. Because web apps cannot spawn local subprocesses, they cannot communicate directly with MCP servers that speak JSON-RPC over stdin/stdout. BridgeKit sits in the middle, translating HTTP POST requests from any web client into stdio subprocess calls, managing per-user session pools of up to 100 concurrent connections, and automatically queuing tool calls that exceed serverless timeout limits (e.g., Vercel's 30-second limit) as background Redis/RQ jobs. A built-in live dashboard provides visibility into sessions, background jobs, and available tools.

Prerequisites

  • Python 3.11 or higher
  • Redis (for session storage and background job queuing — required for production use)
  • Docker and Docker Compose (recommended for the easiest setup)
  • At least one MCP stdio server you want to expose over HTTP
  • Optional: a Vercel account for one-click serverless deployment
1

Clone the repository

Clone MCP BridgeKit to your server or development machine.

git clone https://github.com/mkbhardwas12/mcp-bridgekit.git
cd mcp-bridgekit
2

Start with Docker Compose (recommended)

Docker Compose starts both the BridgeKit FastAPI app and a Redis instance together. This is the fastest way to get a running environment.

docker-compose up
3

Configure the Redis URL

BridgeKit requires a Redis connection for session pooling and background jobs. Set the MCP_BRIDGEKIT_REDIS_URL environment variable to your Redis instance URL.

# In .env or docker-compose.yml:
MCP_BRIDGEKIT_REDIS_URL=redis://localhost:6379/0
4

Connect your web app to BridgeKit

From any HTTP client (React app, curl, n8n, Zapier), send a POST to /chat with the tool name, arguments, and a user ID. BridgeKit manages an isolated MCP server session per user.

curl -X POST http://localhost:8000/chat \
  -H 'Content-Type: application/json' \
  -d '{"user_id": "user-1", "tool_name": "search_docs", "tool_args": {"query": "refund policy"}}'
5

Handle long-running tool calls via background jobs

If a tool call exceeds 25 seconds, BridgeKit queues it automatically and returns a job_id. Poll GET /job/{job_id} to retrieve the result when it completes.

# Check job status:
curl http://localhost:8000/job/abc-123

# Discover available tools for a user session:
curl http://localhost:8000/tools/user-1
6

Deploy to Vercel for serverless hosting (optional)

Click the Deploy with Vercel button in the README or use the Vercel CLI. Set MCP_BRIDGEKIT_REDIS_URL to a hosted Redis (e.g. Upstash) in the Vercel environment variables panel.

MCP BridgeKit Examples

Client configuration

Example HTTP API call from a web application to BridgeKit. The mcp_config block tells BridgeKit which MCP stdio server to launch for this user's session.

{
  "user_id": "user-123",
  "tool_name": "query_db",
  "tool_args": {"sql": "SELECT count(*) FROM users"},
  "mcp_config": {
    "command": "python",
    "args": ["my_mcp_server.py"]
  }
}

Prompts to try

HTTP API interactions with MCP BridgeKit for different use cases.

- POST /chat with {tool_name: 'search_docs', tool_args: {query: 'return policy'}} to call an MCP tool
- GET /tools/user-1 to discover which tools are available in the user's current session
- GET /job/abc-123 to poll for the result of a long-running background job
- DELETE /session/user-1 to manually clean up a user's MCP server session

Troubleshooting MCP BridgeKit

Tool calls time out and job_id is returned but GET /job/{id} never shows 'completed'

Ensure a Redis worker process is running. Background jobs are processed by RQ workers — start one with: rq worker --url redis://localhost:6379/0 in a separate terminal or container.

Session pool is exhausted — new requests are rejected

The default pool limit is 100 concurrent sessions. Idle sessions are auto-evicted by TTL, but you can also manually clean up with DELETE /session/{user_id}. For higher concurrency, deploy multiple BridgeKit instances behind a load balancer.

The MCP stdio server fails to start inside BridgeKit

Verify the command and args in your mcp_config are valid and that the MCP server binary/script is accessible from BridgeKit's working directory or is on the PATH. Test the server independently first with: echo '{}' | your-mcp-command

Frequently Asked Questions about MCP BridgeKit

What is MCP BridgeKit?

MCP BridgeKit is a Model Context Protocol (MCP) server that embeddable mcp stdio → http bridge with background jobs & live dashboard. survives vercel/cloudflare 30s timeouts. now scales to 100+ users. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install MCP BridgeKit?

Follow the installation instructions on the MCP BridgeKit GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with MCP BridgeKit?

MCP BridgeKit works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is MCP BridgeKit free to use?

Yes, MCP BridgeKit is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Developer Tools MCP Servers

Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "mcp-bridgekit": { "command": "npx", "args": ["-y", "mcp-bridgekit"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use MCP BridgeKit?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides