chatgpt-mcp-bridge

chatgpt-mcp-bridge

Enables ChatGPT to perform read-only lookups on public data sources including WHOIS domain information, UK Companies House records, and SEC EDGAR filings through a self-hosted MCP server with an ngrok tunnel.

Category
访问服务器

README

chatgpt-mcp-bridge

A small self-hosted MCP server that gives ChatGPT (Developer Mode / custom connectors) read-only access to a few free public data sources:

  • WHOIS domain lookups (whois_domain)
  • UK Companies House company search and profile lookup (companies_house_search, companies_house_company)
  • SEC EDGAR US public filings, proxied through this same server from sec-edgar-mcp

Wikidata is not hosted here — it's already public at https://wd-mcp.wmcloud.org/mcp/, so just add that URL directly as a fourth ChatGPT connector.

Architecture

ChatGPT ──HTTPS (ngrok tunnel)──> bridge (Express, port 3939)
                                    ├── /mcp           whois + Companies House tools
                                    └── /sec-edgar/mcp ──proxy──> sec-edgar container (port 9870)

One public hostname serves both tool sets — a free ngrok account only grants one static domain, so SEC EDGAR is reverse-proxied under a sub-path instead of getting its own tunnel.

Built on the current official MCP TypeScript SDK pattern: @modelcontextprotocol/server + @modelcontextprotocol/express + @modelcontextprotocol/node.

Setup

1. Environment

cp .env.example .env

Fill in:

  • COMPANIES_HOUSE_API_KEY — free key from developer.company-information.service.gov.uk
  • SEC_EDGAR_USER_AGENT — your real name + email, e.g. Jane Doe (jane@example.com) (SEC requires a genuine identifying user-agent)
  • NGROK_AUTHTOKEN — from your ngrok dashboard
  • PUBLIC_HOSTNAME — your ngrok static domain (free accounts get exactly one, claimed once via the dashboard)

2. Run everything

docker compose up -d

Starts three containers: the bridge, the SEC EDGAR MCP server, and an ngrok tunnel pointing at the bridge.

3. Verify

docker compose logs -f
curl -s -X POST https://<your-domain>.ngrok-free.dev/mcp \
  -H 'Content-Type: application/json' -H 'Accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

4. Register in ChatGPT

Settings → Apps → Advanced settings → enable Developer mode → add connectors (Streamable HTTP, no auth):

  • https://<your-domain>.ngrok-free.dev/mcp
  • https://<your-domain>.ngrok-free.dev/sec-edgar/mcp
  • https://wd-mcp.wmcloud.org/mcp/

Security notes

  • No-auth public endpoint: anyone holding the tunnel URL can call these tools while it's up. All tools here are read-only, but keep the URL private — don't post it anywhere public.
  • createMcpExpressApp's built-in DNS-rebinding protection is configured via PUBLIC_HOSTNAME in allowedHosts/allowedOrigins — requests with an unrecognized Host header get 403.
  • Keep COMPANIES_HOUSE_API_KEY and NGROK_AUTHTOKEN in .env only; never commit it (already gitignored).

Local dev (without Docker)

npm install
npm start          # bridge on :3939

SEC EDGAR needs to be running separately on :9870 for the /sec-edgar/mcp proxy to work — see docker-compose.yml for the exact command, or run the container manually:

docker run -i --rm -p 9870:9870 \
  -e SEC_EDGAR_USER_AGENT="Your Name (your-email@domain.com)" \
  stefanoamorelli/sec-edgar-mcp:latest \
  python -m sec_edgar_mcp.server --transport streamable-http --port 9870

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
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
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
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