Agent skill

mcp

MCP（Model Context Protocol）集成技能，支持自动安装、配置和使用MCP服务器，连接各种数据源和工具。当用户需要让AI模型安全访问外部数据源、数据库、API或文件系统时使用此技能。

View SKILL.md on GitHub Repository
Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/mcp-chengyihua-nanobot-ts

Metadata

Additional technical details for this skill

nanobot 
{
    "emoji": "\ud83d\udd0c",
    "install": [
        {
            "id": "mcp-server",
            "kind": "command",
            "label": "Install MCP Servers",
            "command": "npm install -g @modelcontextprotocol/server @modelcontextprotocol/server-filesystem @modelcontextprotocol/server-sqlite @modelcontextprotocol/server-http"
        }
    ],
    "requires": {
        "bins": [
            "node",
            "npm"
        ]
    }
}

SKILL.md

MCP（Model Context Protocol）集成技能

什么是MCP？

MCP是Anthropic推出的协议，允许AI模型安全地访问外部工具和数据源，而无需直接集成到模型中。

核心优势：

  1. 安全性：模型不能直接执行代码，只能通过MCP服务器访问
  2. 灵活性：可以连接各种数据源（数据库、API、文件系统等）
  3. 标准化：统一的协议，不同模型都可以使用

快速开始

1. 安装MCP服务器

bash
# 安装常用MCP工具
npm install -g @modelcontextprotocol/server-filesystem @modelcontextprotocol/server-sqlite @modelcontextprotocol/server-http

2. 启动MCP服务器

bash
# 启动文件系统服务器
npx @modelcontextprotocol/server-filesystem --directory /path/to/data

# 启动SQLite服务器
npx @modelcontextprotocol/server-sqlite --database /path/to/database.db

# 启动HTTP服务器
npx @modelcontextprotocol/server-http --port 8080

3. 在nanobot中配置MCP

javascript
// 在nanobot配置中添加MCP支持
const { createMCPClient } = require('@ai-sdk/openai');

const mcpClient = createMCPClient({
  transport: 'stdio',
  command: 'mcp-server',
  args: ['filesystem', '--directory', '/path/to/data']
});

常用MCP服务器

1. 文件系统服务器

bash
# 访问本地文件系统
mcp-server filesystem --directory /Users/chengyihua/Documents

# 使用示例：读取文件列表
curl -X POST http://localhost:8080/tools/list \
  -H "Content-Type: application/json" \
  -d '{"path": "/Users/chengyihua/Documents"}'

2. SQLite服务器

bash
# 连接SQLite数据库
mcp-server sqlite --database /path/to/database.db

# 使用示例：执行SQL查询
curl -X POST http://localhost:8080/tools/query \
  -H "Content-Type: application/json" \
  -d '{"sql": "SELECT * FROM users LIMIT 10"}'

3. HTTP服务器

bash
# 启动HTTP API服务器
mcp-server http --port 8080

# 使用示例：调用API
curl -X POST http://localhost:8080/tools/fetch \
  -H "Content-Type: application/json" \
  -d '{"url": "https://api.example.com/data", "method": "GET"}'

4. PostgreSQL服务器

bash
# 连接PostgreSQL数据库
mcp-server postgres --connection-string "postgresql://user:password@localhost:5432/dbname"

5. GitHub服务器

bash
# 访问GitHub API
mcp-server github --token YOUR_GITHUB_TOKEN

在nanobot中使用MCP

1. 配置MCP客户端

javascript
// 在nanobot的agent配置中添加
const mcpConfig = {
  servers: [
    {
      name: 'filesystem',
      transport: 'stdio',
      command: 'mcp-server',
      args: ['filesystem', '--directory', process.cwd()]
    },
    {
      name: 'sqlite',
      transport: 'stdio', 
      command: 'mcp-server',
      args: ['sqlite', '--database', '/path/to/data.db']
    }
  ]
};

2. 使用MCP工具

javascript
// 通过MCP访问文件系统
const files = await mcpClient.callTool('filesystem.list', {
  path: '/Users/chengyihua/Documents'
});

// 通过MCP执行SQL查询
const results = await mcpClient.callTool('sqlite.query', {
  sql: 'SELECT * FROM products WHERE price > 100'
});

// 通过MCP调用HTTP API
const response = await mcpClient.callTool('http.fetch', {
  url: 'https://api.example.com/data',
  method: 'GET'
});

自动安装脚本

安装脚本 (scripts/install_mcp.sh)

bash
#!/bin/bash

# MCP自动安装脚本
echo "开始安装MCP..."

# 检查Node.js
if ! command -v node &> /dev/null; then
    echo "安装Node.js..."
    brew install node
fi

# 安装MCP服务器
echo "安装MCP服务器..."
npm install -g @modelcontextprotocol/server

# 安装常用MCP工具
echo "安装MCP工具..."
npm install -g @modelcontextprotocol/server-filesystem
npm install -g @modelcontextprotocol/server-sqlite
npm install -g @modelcontextprotocol/server-http
npm install -g @modelcontextprotocol/server-postgres
npm install -g @modelcontextprotocol/server-github

# 验证安装
echo "验证安装..."
mcp-server --version

echo "MCP安装完成！"

启动脚本 (scripts/start_mcp.sh)

bash
#!/bin/bash

# 启动MCP服务器
echo "启动MCP服务器..."

# 启动文件系统服务器
mcp-server filesystem --directory /Users/chengyihua/Documents &

# 启动SQLite服务器（如果有数据库）
if [ -f "/path/to/database.db" ]; then
    mcp-server sqlite --database /path/to/database.db &
fi

# 启动HTTP服务器
mcp-server http --port 8080 &

echo "MCP服务器已启动"
echo "文件系统服务器: 访问本地文件"
echo "HTTP服务器: http://localhost:8080"

使用示例

示例1：通过MCP读取文件

bash
# 使用curl调用MCP服务器
curl -X POST http://localhost:8080/tools/read \
  -H "Content-Type: application/json" \
  -d '{"path": "/Users/chengyihua/Documents/example.txt"}'

示例2：通过MCP查询数据库

bash
# 查询SQLite数据库
curl -X POST http://localhost:8080/tools/query \
  -H "Content-Type: application/json" \
  -d '{"sql": "SELECT name, price FROM products ORDER BY price DESC LIMIT 5"}'

示例3：通过MCP调用GitHub API

bash
# 获取GitHub仓库信息
curl -X POST http://localhost:8080/tools/github \
  -H "Content-Type: application/json" \
  -d '{"endpoint": "/repos/owner/repo", "method": "GET"}'

故障排除

常见问题

1. MCP服务器无法启动

bash
# 检查端口占用
lsof -i :8080

# 检查MCP安装
mcp-server --version

2. 连接被拒绝

bash
# 检查防火墙设置
sudo ufw status

# 检查MCP服务器日志
tail -f /var/log/mcp-server.log

3. 权限问题

bash
# 检查文件权限
ls -la /path/to/data

# 以正确用户运行
sudo -u username mcp-server filesystem --directory /path/to/data

调试模式

bash
# 启用调试日志
DEBUG=mcp:* mcp-server filesystem --directory /path/to/data

# 详细日志
mcp-server filesystem --directory /path/to/data --verbose

安全最佳实践

1. 访问控制

bash
# 使用认证令牌
mcp-server filesystem --directory /path/to/data --token SECRET_TOKEN

# 限制访问IP
mcp-server http --port 8080 --allowed-ips 192.168.1.0/24

2. 数据加密

bash
# 使用HTTPS
mcp-server http --port 8443 --ssl-cert /path/to/cert.pem --ssl-key /path/to/key.pem

3. 审计日志

bash
# 启用访问日志
mcp-server filesystem --directory /path/to/data --log-file /var/log/mcp-access.log

高级配置

1. 多服务器配置

json
{
  "servers": [
    {
      "name": "production-db",
      "transport": "stdio",
      "command": "mcp-server",
      "args": ["postgres", "--connection-string", "postgresql://user:pass@prod-db:5432/prod"]
    },
    {
      "name": "development-fs",
      "transport": "stdio",
      "command": "mcp-server", 
      "args": ["filesystem", "--directory", "/home/dev/data"]
    }
  ]
}

2. 负载均衡

bash
# 启动多个MCP实例
mcp-server http --port 8081 &
mcp-server http --port 8082 &
mcp-server http --port 8083 &

# 使用负载均衡器
nginx -c /path/to/nginx-mcp.conf

3. 监控和指标

bash
# 启用Prometheus指标
mcp-server http --port 8080 --metrics-port 9090

# 查看健康检查
curl http://localhost:8080/health

参考资料

官方文档

社区资源

相关工具

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

majiayu000/claude-skill-registry

agent-ops-spec

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-state

Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-spec

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-testing

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-testing

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-state

Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.

163 31
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results