Om Sacred Space — ChatGPT Healing Connector
Enables ChatGPT to answer questions about sound healing, self-healing, energy work, and Om Sacred Space's offerings through a minimal MCP server.
README
Om Sacred Space — ChatGPT Healing Connector
A minimal MCP server that lets ChatGPT answer
questions about sound healing, self-healing, energy work, and Om Sacred Space's
offerings. This is the fast Q&A prototype — content lives in knowledge.js and
can be expanded any time.
What's inside
| File | Purpose |
|---|---|
server.js |
MCP server over streamable HTTP, exposes 3 tools |
knowledge.js |
The healing Q&A knowledge base (~10 topics from the live site) |
test-client.js |
Smoke test that calls every tool without needing ChatGPT |
Tools
ask_healing— direct natural-language Q&A (the main one).search— find matching topics (ChatGPT / Deep Research compatible shape).fetch— return the full answer for a topic id fromsearch.
All tools are read-only. Every answer carries a wellness disclaimer.
Run locally
cd chatgpt-connector
npm install
npm start # serves http://localhost:8787/mcp
In another terminal, smoke-test it:
node test-client.js
Try it in ChatGPT (Developer Mode)
- Tunnel the local server:
ngrok http 8787 - In ChatGPT: Settings → Connectors → Advanced → enable Developer mode (requires a paid plan).
- Create a connector and paste the tunnel URL with
/mcp, e.g.https://<subdomain>.ngrok.app/mcp. Auth: None for testing. - Start a chat, attach the connector, and ask a healing question.
Deploy for real (public availability)
For submission to the ChatGPT Apps Directory the endpoint must be a public
HTTPS host (no localhost/ngrok). Good options: Vercel, Fly.io, Render, Railway.
Set a Content Security Policy, store any secrets in env vars, and point a
subdomain such as mcp.omsacredspace.com at it.
Expand the knowledge base
Add objects to the HEALING_QA array in knowledge.js — each needs id,
question, answer, tags (keywords for matching), and url. Restart the
server to pick up changes.
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。