MCP Bridge
这是一个为 Cocos Creator 设计的 MCP (Model Context Protocol) 桥接插件，用于连接外部 AI 工具与 Cocos Creator 编辑器，实现对场景、节点等资源的自动化操作。
What is MCP Bridge?
MCP Bridge is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to 这是一个为 cocos creator 设计的 mcp (model context protocol) 桥接插件，用于连接外部 ai 工具与 cocos creator 编辑器，实现对场景、节点等资源的自动化操作。
这是一个为 Cocos Creator 设计的 MCP (Model Context Protocol) 桥接插件，用于连接外部 AI 工具与 Cocos Creator 编辑器，实现对场景、节点等资源的自动化操作。
This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- 这是一个为 Cocos Creator 设计的 MCP (Model Context Protocol) 桥接插件，用于
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx mcp-bridgeConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use MCP Bridge
MCP Bridge is a Docker-based MCP server that exposes a local llama.cpp instance (or any remote llama.cpp-compatible server) through the Model Context Protocol over HTTP with bearer-token authentication. It runs FastMCP behind a Streamable HTTP transport and forwards chat completions to llama.cpp's OpenAI-compatible API, letting MCP clients like Claude Code treat a self-hosted language model as a first-class tool provider. This is useful for teams running air-gapped models or wanting to add private LLM inference as a capability inside their AI assistant workflows.
Prerequisites
- Docker and Docker Compose installed
- A running llama.cpp server (local or remote) with its OpenAI-compatible API accessible
- openssl available to generate a secure API token
- An MCP-compatible client that supports HTTP transport (e.g., Claude Code CLI)
Clone the repository
Download the project source which includes the Dockerfile, docker-compose.yml, and .env.example configuration template.
git clone https://github.com/loopsketch/mcp-bridge.git
cd mcp-bridgeCreate and configure the .env file
Copy the example environment file and set a secure random API token. Optionally point LLAMA_CPP_SERVER_URL to a remote llama.cpp instance if you are not running one locally.
cp .env.example .env
echo "MCP_API_TOKEN=$(openssl rand -hex 32)" >> .envReview key environment variables
The .env file controls transport, host, port, and llama.cpp connection settings. Edit as needed before starting the service.
# Key variables in .env:
# MCP_TRANSPORT=streamable-http
# FASTMCP_HOST=0.0.0.0
# FASTMCP_PORT=8000
# LLAMA_CPP_SERVER_URL=http://localhost:8080
# LLAMA_CPP_SERVER_MODEL=your-model-name
# MCP_API_TOKEN=<generated above>
# MAX_INPUT_CHARS=200000Start the service
Build and launch the Docker container in detached mode. The MCP endpoint will be available at http://localhost:8000/mcp.
docker compose up -d --buildVerify the endpoint
Check that the server is running and authentication is required. A 406 response to an authenticated request is expected behavior indicating the server is online.
curl -i -H "Authorization: Bearer $MCP_API_TOKEN" http://localhost:8000/mcpRegister with your MCP client
Add the HTTP-transport MCP server to Claude Code or another MCP-compatible client using the generated token.
claude mcp add --transport http local-agent-helper \
http://localhost:8000/mcp \
--header "Authorization: Bearer $MCP_API_TOKEN"MCP Bridge Examples
Client configuration
HTTP-transport MCP servers require a URL and auth header rather than a command/args block. Use the Claude Code CLI to register or add the entry manually to your config.
# Claude Code CLI registration:
claude mcp add --transport http local-agent-helper \
http://localhost:8000/mcp \
--header "Authorization: Bearer YOUR_MCP_API_TOKEN"Prompts to try
Once connected, the local llama.cpp model is available as a tool provider through your MCP client.
- "Use the local model to summarize this document"
- "Ask the local-agent-helper to generate a Python function for binary search"
- "What models are currently available through the local llama.cpp server?"Troubleshooting MCP Bridge
Docker container starts but returns connection refused when calling the MCP endpoint
Ensure FASTMCP_HOST is set to 0.0.0.0 (not 127.0.0.1) in the .env file so the server listens on all interfaces inside the container, and that port 8000 is not blocked by a local firewall.
Requests to llama.cpp fail with a timeout or connection error
Verify LLAMA_CPP_SERVER_URL points to the correct address reachable from inside Docker. If llama.cpp is running on the host machine, use 'http://host.docker.internal:8080' instead of 'http://localhost:8080'.
Authentication fails with 401 Unauthorized
Make sure MCP_API_TOKEN in the .env file matches the token passed in the Authorization header. The token must be at least 32 characters. Rebuild the container after changing .env with 'docker compose up -d --build'.
Frequently Asked Questions about MCP Bridge
What is MCP Bridge?
MCP Bridge is a Model Context Protocol (MCP) server that 这是一个为 cocos creator 设计的 mcp (model context protocol) 桥接插件，用于连接外部 ai 工具与 cocos creator 编辑器，实现对场景、节点等资源的自动化操作。 It connects AI assistants to external tools and data sources through a standardized interface.
How do I install MCP Bridge?
Follow the installation instructions on the MCP Bridge GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with MCP Bridge?
MCP Bridge works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is MCP Bridge free to use?
Yes, MCP Bridge is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
MCP Bridge Alternatives — Similar Developer Tools Servers
Looking for alternatives to MCP Bridge? Here are other popular developer tools servers you can use with Claude, Cursor, and VS Code.
Ecc
★ 188.2kThe agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
Javaguide
★ 155.8kJava 面试 & 后端通用面试指南，覆盖计算机基础、数据库、分布式、高并发、系统设计与 AI 应用开发
Gemini CLI
★ 104.5kA secure MCP server that wraps the Google Gemini CLI, allowing clients to query Gemini models using local OAuth sessions without requiring an API key. It provides tools for model interaction and diagnostics with built-in protection against command in
Awesome MCP Servers
★ 87.3k⭐ Curated list of Model Context Protocol (MCP) servers - tools that extend Claude Desktop, Cursor, Windsurf, and other MCP clients with custom capabilities.
MCP Servers
★ 86.0kModel Context Protocol Servers
CC Switch
★ 77.5kA cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Gemini CLI & Hermes Agent. Only official website: ccswitch.io
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.
Set Up MCP Bridge in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use MCP Bridge?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.