Crawl MCP

Name: Crawl MCP
Author: gqy20

gqy20

基于 crawl4ai 做单页、整站与批量网页爬取，并搜索文本、新闻、图片、图书和视频，可选 LLM 后处理。

接入配置

选择客户端，复制配置，再按项目文档补齐环境变量。

claude mcp add crawl -- uvx crawl-mcp

~/.codex/config.toml

[mcp_servers.crawl]
command = "uvx"
args = ["crawl-mcp"]

# codex mcp add crawl -- uvx crawl-mcp

~/.cursor/mcp.json

{
  "mcpServers": {
    "crawl": {
      "command": "uvx",
      "args": [
        "crawl-mcp"
      ]
    }
  }
}

.vscode/mcp.json

{
  "servers": {
    "crawl": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "crawl-mcp"
      ]
    }
  }
}

~/.gemini/settings.json

{
  "mcpServers": {
    "crawl": {
      "command": "uvx",
      "args": [
        "crawl-mcp"
      ]
    }
  }
}

claude_desktop_config.json

{
  "mcpServers": {
    "crawl": {
      "command": "uvx",
      "args": [
        "crawl-mcp"
      ]
    }
  }
}

.trae/mcp.json

{
  "mcpServers": {
    "crawl": {
      "command": "uvx",
      "args": [
        "crawl-mcp"
      ]
    }
  }
}