BridgeDesk

BridgeDesk

A desktop launcher and local MCP workspace server that enables ChatGPT and other MCP clients to securely read, edit, search, run commands, and show changes in selected local project folders.

Category
访问服务器

README

BridgeDesk

About

BridgeDesk is a desktop launcher and local MCP workspace server for ChatGPT and other MCP clients. It lets a trusted MCP client work with selected local project folders through explicit tools for reading, editing, searching, running commands, and showing changes.

BridgeDesk is Windows-first. It is designed for local development workflows where you choose one approved project folder, start the BridgeDesk server, and connect an MCP-capable client through the generated MCP URL.

Download BridgeDesk

For normal Windows use, download the latest installer from GitHub Releases:

https://github.com/stanic-mah/bridgedesk/releases/latest/download/BridgeDesk-Setup.exe

The installer build supports BridgeDesk's in-app update checks. A portable exe is also attached to each release for manual use, but the installer is the recommended download.

BridgeDesk has two entrypoints:

  • an Electron desktop launcher for setup, tunnels, process control, and copyable connection details
  • a CLI and MCP server for headless or scripted use

Requirements

  • Node >=20.12 <27
  • npm
  • Git
  • Bash-compatible shell; Git Bash is the simplest Windows option
  • Optional: cloudflared for the launcher-managed quick tunnel

Desktop Launcher

Install dependencies and start the launcher from a local checkout:

npm install
npm run desktop

The launcher checks local requirements, lets you choose one project folder, starts a Cloudflare tunnel when available, starts the BridgeDesk server, and copies the MCP URL and Owner password for the MCP client.

BridgeDesk supports two tunnel modes:

  • Quick Tunnel: easiest setup. Cloudflare gives BridgeDesk a temporary trycloudflare.com URL, and that URL changes when the tunnel is restarted.
  • Permanent Tunnel: best for regular use. BridgeDesk uses your Cloudflare account, domain, and named tunnel so the MCP URL can stay fixed, for example https://mcp.yourdomain.com/mcp.

Use Permanent Tunnel if you do not want to re-add the MCP server in ChatGPT after closing the app or restarting the computer.

If Cloudflare Tunnel is not installed, paste any HTTPS tunnel URL into the launcher. The URL should point to:

http://127.0.0.1:7676

Use this MCP endpoint in the client:

https://your-tunnel-host.example.com/mcp

CLI

Build and run locally:

npm install
npm run build
node dist/cli.js init
node dist/cli.js serve

After publishing or linking the package, the public command is:

bridgedesk init
bridgedesk serve
bridgedesk doctor
bridgedesk config get
bridgedesk config set publicBaseUrl https://your-tunnel-host.example.com

The default local endpoint is:

http://127.0.0.1:7676/mcp

Configuration

BridgeDesk stores local configuration in:

~/.bridgedesk/config.json
~/.bridgedesk/auth.json

Important environment variables:

  • BRIDGEDESK_PUBLIC_BASE_URL
  • BRIDGEDESK_ALLOWED_ROOTS
  • BRIDGEDESK_OAUTH_OWNER_TOKEN
  • BRIDGEDESK_STATE_DIR
  • BRIDGEDESK_WORKTREE_ROOT
  • BRIDGEDESK_TOOL_MODE

See docs/configuration.md for the full reference.

Security

BridgeDesk is powerful local access to selected folders on your machine. Keep allowed roots narrow, protect auth.json, and connect only trusted MCP clients. The desktop launcher rejects drive roots and the whole user folder as project roots.

See docs/security.md for the full model.

Development

npm install
npm run typecheck
npm test
npm run build
npm run desktop

The build produces:

  • dist/cli.js for the CLI
  • dist/server.js for the MCP server
  • dist/ui for MCP app widgets
  • dist/desktop for the Electron launcher

License

BridgeDesk is distributed under the MIT License. See LICENSE.

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选