mcp-quickserver

mcp-quickserver

A working MCP server template for notes management (add, get, search, delete) with in-memory store, demonstrating tools, resources, and prompts for easy forking and customization.

Category
访问服务器

README

CI

mcp-quickserver

A working MCP (Model Context Protocol) server template. Fork it, swap the store, add your tools.

Demonstrates all three MCP primitives — tools, resources, and prompts — with both stdio and SSE transports, proper error handling, and a test harness that runs without a connected client.


What's included

Primitive What it does
Tools add_note, get_note, search_notes, delete_note
Resources notes://all (full list), notes://{id} (single note)
Prompts summarize_notes — builds a summarisation prompt, optionally filtered by tag

The backing store is an in-memory NoteStore. Replacing it with SQLite, Postgres, or Redis is a one-file change.


Install

pip install -e .

Requires Python ≥ 3.10 and the mcp SDK.


Wire to Claude Desktop

Copy claude_desktop_config.json.example into your Claude Desktop config:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "quickserver": {
      "command": "python",
      "args": ["-m", "mcpquickserver"],
      "cwd": "/absolute/path/to/mcp-quickserver"
    }
  }
}

Restart Claude Desktop. The tools appear automatically in the tool picker.

Wire to Claude Code

# From within your project
claude mcp add quickserver -- python -m mcpquickserver

Or add to .claude/settings.json:

{
  "mcpServers": {
    "quickserver": {
      "command": "python",
      "args": ["-m", "mcpquickserver"]
    }
  }
}

Run as an HTTP server (SSE transport)

MCP_HOST=0.0.0.0 MCP_PORT=9000 python -m mcpquickserver --transport sse

SSE transport is useful when the client runs in a different process or on a different machine.


Code layout

src/mcpquickserver/
├── server.py    ← tool/resource/prompt registration against a NoteStore
├── store.py     ← NoteStore business logic (swap this for persistence)
└── __main__.py  ← CLI entry point

Adding a tool

# server.py
@mcp.tool()
def word_count(text: str) -> str:
    """Count the words in a block of text."""
    return str(len(text.split()))

FastMCP generates the JSON schema from the type annotations. The docstring becomes the tool description shown to the model.

Adding a resource

@mcp.resource("config://settings")
def settings_resource() -> str:
    """Current server configuration as JSON."""
    return json.dumps({"version": "0.1.0", "store": "memory"})

Dynamic URI segments work too — notes://{note_id} captures note_id from the URI and passes it as a parameter.

Adding a prompt template

@mcp.prompt()
def draft_email(recipient: str, topic: str) -> str:
    """Build a prompt to draft a professional email."""
    return f"Please draft a professional email to {recipient} about: {topic}"

Error handling

Tools that raise an exception return an is_error: true result to the model so it can recover — the exception does not propagate to the caller or crash the server:

@mcp.tool()
def get_note(note_id: str) -> str:
    """Retrieve a note by its ID."""
    note = store.get(note_id)
    if note is None:
        raise ValueError(f"No note with id {note_id!r}")  # → is_error result
    return json.dumps(note, indent=2)

Tests

pip install -e ".[dev]"
pytest -v

Tests cover the NoteStore business logic and all tool/resource/prompt handlers directly. No connected client required. The store module attribute is monkeypatched per test so each test starts with a clean slate.


Replacing the store

NoteStore in store.py exposes five methods: add, get, delete, search, all. Any object that implements those five is a drop-in replacement:

# e.g. SQLite backend
class SqliteNoteStore:
    def __init__(self, db_path: str): ...
    def add(self, title, content, tags=None) -> str: ...
    def get(self, note_id) -> dict | None: ...
    def delete(self, note_id) -> bool: ...
    def search(self, query="", tag="") -> list[dict]: ...
    def all(self) -> list[dict]: ...

Swap the store = NoteStore() line in server.py and the rest is unchanged.

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