Shanhai MCP Server

Shanhai MCP Server

An out-of-the-box multi-tenant MCP server that lets MCP clients access knowledge retrieval, task dispatch, deliverable management, and multi-agent consultation services via SSE or stdio.

Category
Visit Server

README

Shanhai MCP Server

一个开箱即用的多租户 MCP Server:让 WorkBuddy / Claude / 任意 MCP 客户端通过 SSE 或 stdio 接入你的服务能力——知识检索、任务投递、交付物管理、多智能体咨询。

License: MIT


特性

工具 能力
list_capabilities 平台能力、权限边界与信任背书声明
search_study 知识库 / 学习模块检索(RAG 语义 + 关键词兜底)
dispatch_task 提交协作任务(内容安检 → 投递到后端场景编排引擎)
get_task_status 查询任务状态(异步任务轮询)
submit_result 提交任务执行结果(同样过内容安检)
list_deliverables 列出本租户交付目录文件(顶层 + 白名单子目录)
get_deliverable 拉取交付文件内容(水印 + 每日限读防扩散)
ask_think_tank 咨询台:复杂问题 → 多角色独立评审 → 结构化方案

设计目标

  • 🔒 fail-closed:无 Key 拒绝一切工具调用,绝不静默降级
  • 🏢 多租户隔离:每租户独立交付目录 / 任务态 / 搜索根(进程级 + Key 级双重隔离)
  • 🛡️ 内容安检:所有用户输入过内容安检层(prompt 注入 / 敏感词),deny 即拒
  • 📊 配额 + 熔断:每租户每日/每月配额,异常调用熔断防滥用
  • 🧱 异步协作:长任务(多智能体评审等)异步受理,task_id 轮询取结果

快速开始

1. 安装

git clone <your-repo-url> && cd shanhai-mcp-server
python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

2. 配置

复制环境变量模板并填写:

cp .env.example .env
# 编辑 .env:SHANHAI_API_KEY 必填(fail-closed 主 Key)

3. 运行(stdio 模式,本地连接器)

python3 server.py

4. 运行(SSE 模式,远程服务)

cp start-sse-public.example.sh start-sse-public.sh
# 按注释配置多租户环境变量
./start-sse-public.sh

SSE 端点:https://your-domain.com/mcp-public/sse

5. 客户端接入(WorkBuddy / 任意 MCP 客户端)

远程 SSE 服务正确写法("type": "sse",不是 "type": "http" / "transport": "sse"):

{
  "mcpServers": {
    "shanhai-city": {
      "type": "sse",
      "url": "https://your-domain.com/mcp-public/sse",
      "headers": {
        "Authorization": "Bearer your-mcp-key-here"
      }
    }
  }
}

💡 本地 stdio 连接器用 "type": "stdio"。配置完成后在客户端「连接器管理」里点 信任/启用 激活。


多租户隔离(SSE 客户模式)

SHANHAI_TASK_DIR        # 任务态目录(与内部物理隔离,缺省拒绝启动)
SHANHAI_CUSTOMER_DIR    # 默认租户交付目录
SHANHAI_CUSTOMER_DIR_MAP  # 按 Key 映射交付目录(key1:dir1,key2:dir2)
SHANHAI_DELIVERABLE_SUBDIRS  # 交付子目录白名单("知识卡片,咨询方案")
SHANHAI_WISDOM_ROOT     # 租户搜索根(脱敏学习模块)
  • 每租户独立 Key → 独立交付目录,交叉访问被拦(forbidden / file_not_found
  • 文件名白名单正则 + resolve() 防路径穿越
  • 客户模式不返回服务器内部路径(result_path 恒为空,交付物走 get_deliverable

安全设计

机制
认证 Bearer Key 白名单(SHANHAI_MCP_KEYS)+ fail-closed
隔离 进程级(内部/客户双实例)+ 目录锁死 + 搜索根隔离
安检 内容安检层(注入检测 / 敏感词),deny 即拒
防滥用 交付物每日限读 + 咨询台月配额 + 熔断(1h 超阈值冷静)
审计 工具调用统计落盘(SHANHAI_USAGE_LOG_DIR

环境变量

完整变量表见 .env.example(所有密钥/路径均为占位符)。

项目结构

shanhai-mcp-server/
├── server.py                 # MCP Server 主体(FastMCP,8 工具)
├── start-sse-public.example.sh  # SSE 客户实例启动模板
├── mcp.json.example          # 客户端接入配置模板
├── .env.example              # 环境变量模板
├── requirements.txt
├── LICENSE                   # MIT
└── README.md

免责声明

本项目由山海城邦出品,以 MIT 协议开源,仅供学习与参考。实际部署请结合自身业务环境调整:务必替换所有密钥、配置独立域名与 TLS、按需裁剪工具权限。作者不对因使用本软件造成的任何直接或间接损失承担责任。


Shanhai · 山海城邦出品

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured