AI OS MCP Server

AI OS MCP Server

Exposes a Notion-synced AI OS as read-only MCP tools, enabling MCP-compatible AI clients to search the vault, fetch page contents, list projects, and check project statuses without scraping public links.

Category
访问服务器

README

AI OS MCP Server

Exposes Felix's AI OS (synced in Notion) as MCP tools, so any MCP-compatible AI client can query it directly instead of scraping a public link.

Honest limitation: this was written without the ability to actually npm install or compile-test it (no network access in the environment it was built in). The code follows the current MCP TypeScript SDK and Notion client API shapes correctly as far as known, but run the build step yourself and check for errors before trusting it — don't assume it's flawless just because it was written carefully.

Tools It Exposes

  • search_vault(query) — keyword search across the whole AI OS
  • get_page(page_name) — full content of one named page
  • list_projects() — the 10_Projects index page's current contents
  • get_project_status(project_name) — one project's current status specifically

Setup

1. Create a Notion integration

Go to notion.so/my-integrations → New integration → name it (e.g. "AI OS MCP") → copy the "Internal Integration Secret" (starts with secret_ or ntn_).

2. Share the AI OS with it

Open the AI OS root page in Notion → ... menu → Connections → add the integration you just created. This is separate from "Share to web" — it's what actually lets the integration's API calls see the pages.

3. Set the API key

cp .env.example .env
# paste your key into .env

4. Build and run with Docker

docker compose build
docker compose run --rm ai-os-mcp

This runs the server over stdio — it's meant to be launched by an MCP client (Claude Desktop, etc.), not left running standalone in a terminal long-term.

5. Point an MCP client at it

For Claude Desktop, add to its MCP config (claude_desktop_config.json):

{
  "mcpServers": {
    "ai-os": {
      "command": "docker",
      "args": ["compose", "-f", "/full/path/to/ai-os-mcp/docker-compose.yml", "run", "--rm", "ai-os-mcp"]
    }
  }
}

Things Worth Checking Before Trusting It

  • Does npm install succeed cleanly, or are there version mismatches with the SDK?
  • Does npm run build compile without errors?
  • Does search_vault actually return results — this depends on the integration having been shared with the right pages (step 2).
  • Table content in get_page may render roughly — Notion's block API is nested and this pulls table rows one level deep, not deeply verified.

What This Doesn't Do

No writing/editing — read-only by design, so it can't accidentally modify the vault. No automation trigger — still has to be invoked by an MCP client on request, consistent with the AI OS's "manual, chat-triggered" execution decision.

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