mcp-thread-persistence-demo
MCP 线程持久化演示
一个极简的 MCP(Model Context Protocol)服务器,将聊天线程笔记持久化到 SQLite,并附带一个真实客户端,通过实际的 stdio 传输来调用它(不仅仅是直接调用底层函数)。
作为学习练习构建,模仿真实的自由职业工作模式:"修复/实现 MCP 工具调用,将聊天线程状态持久化到数据库,采用 upsert 语义(新建则创建,已存在则更新,每个线程不产生重复行)。"
这里有什么
server.py— 暴露三个工具的 MCP 服务器:save_note(user_id, thread_id, content)— 以(user_id, thread_id)为键 upsert 到 SQLite。对同一线程的第二次调用会更新现有行,而不是创建重复行。get_note(user_id, thread_id)— 读回已保存的笔记。list_threads(user_id)— 列出某个用户的所有线程。
test_client.py— 一个真实的 MCP 客户端,通过 stdio 连接到server.py,调用list_tools,然后按顺序执行全部三个工具(包括对同一线程调用两次save_note,以验证 upsert 不会创建重复行)。
Related MCP server: tether
运行
python3.12 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python test_client.py备注
特意锁定在 mcp<2:mcp SDK 的 2.x 版本将 FastMCP 重命名为 MCPServer,并更改了其他 API。目前市面上大多数现有教程/代码库仍以 1.x 的 FastMCP API 为目标,因此本演示也以该版本为目标。
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
Persistent memory layer for AI tools. Save and recall notes across Claude and other MCP clients.
MCP-native notes and memory for ChatGPT, Claude, and other AI tools.
Google Keep-style notes app with an MCP server for AI agents to read/write notes.
Persistent memory shared across Claude, ChatGPT, Grok and other MCP clients.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables adding, getting, and listing notes through MCP, with state persisted to a JSON file in the home directory.MIT
- AlicenseAqualityBmaintenanceEnables MCP-compatible agents to remember, recall, link, and forget durable notes across devices, with local SQLite storage and optional libSQL/Turso sync.438 PyPI1MIT
- AlicenseNot gradedqualityCmaintenanceEnables creating, retrieving, updating, deleting, and listing notes through a simple MCP interface, with configurable note storage.MIT
- FlicenseNot gradedqualityAmaintenanceEnables MCP-compatible clients such as Claude to create, link, search, explore, and synthesize atomic notes using a Markdown source of truth with an SQLite index for traversal and querying.-