Note App MCP Server
MCP server for a JWT-authenticated note-taking API, exposing CRUD tools for notes.
README
Note App — FastAPI + MCP Demo
A simple JWT-authenticated note-taking API built with FastAPI, backed by SQLite via SQLAlchemy, and exposed as MCP tools for agents using fastapi-mcp.
Requirements
- Python 3.13+
- Node.js (only needed to run the MCP Inspector via
npx)
Setup
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
Create a .env file in the project root (see src/environment.py for the full list of settings):
USERNAME=admin
PASSWORD=changeme
JWT_SECRET=some-long-random-secret
JWT_EXPIRATION_MINUTES=60 # optional, defaults to 60
Running the server
python server.py
This starts uvicorn on http://0.0.0.0:8000, serving src.app:app.
- Swagger UI: http://localhost:8000/docs
- The root URL
/redirects to/docs
Authentication
Get an access token:
curl -X POST http://localhost:8000/auth/login \
-H "Content-Type: application/json" \
-d '{"username": "admin", "password": "changeme"}'
Use the returned access_token as a Bearer token on all /notes endpoints:
curl http://localhost:8000/notes/ \
-H "Authorization: Bearer <access_token>"
Notes API
| Method | Path | Description |
|---|---|---|
| GET | /notes/ |
List notes |
| GET | /notes/{id} |
Get a note |
| POST | /notes/ |
Create a note |
| PUT | /notes/{id} |
Update a note |
| DELETE | /notes/{id} |
Delete a note |
Using the MCP server
The app mounts an MCP server at /mcp, exposing every endpoint tagged Note as an MCP tool. Requests require the same Bearer token as the REST API, forwarded via the authorization header.
Inspecting it with MCP Inspector
- Start the server:
python server.py - In a separate terminal, launch the inspector:
npx @modelcontextprotocol/inspector - In the inspector UI, set:
- Transport: Streamable HTTP
- URL:
http://127.0.0.1:8000/mcp - Header:
Authorization: Bearer <access_token>(obtained from/auth/login)
- Click Connect, go to the Tools tab, and click List Tools to see
list_notes,get_note,create_note,update_note,delete_note. - Select a tool, fill in parameters, and click Run Tool to test it.
Connecting from an MCP client (e.g. Claude Desktop)
Since the server requires auth headers, use mcp-remote as a bridge:
{
"mcpServers": {
"note-app": {
"command": "npx",
"args": [
"mcp-remote",
"http://localhost:8000/mcp",
"--header",
"Authorization:${AUTH_HEADER}"
],
"env": {
"AUTH_HEADER": "Bearer <access_token>"
}
}
}
}
Project structure
server.py # entrypoint (uvicorn runner)
src/
app.py # FastAPI app + MCP mount
environment.py # settings loaded from .env
modules/
auth/ # login route, JWT issuing/verification
note/ # notes CRUD (model, schema, service, route)
sqlalchemy/ # base model + engine/session setup
Dev tooling
Config files are included for:
ruff check . # linting (ruff.toml)
mypy . # type checking (mypy.ini)
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。