agentcurl

chenhaodev/agentcurl

★ 0 stars Python Browser/Web Updated today

Switchable web-crawler middleware: one interface over pluggable backends (static/trafilatura, Playwright, crawl4ai, Firecrawl, jina) with DeepSeek structured extraction. Repo + MCP server + Claude Code SKILL.

View on GitHub → Try with Claude — $10 free →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "agentcurl": {
      "command": "uvx",
      "args": [
        "agentcurl"
      ]
    }
  }
}

Or install with pip: pip install agentcurl

README Excerpt

> 一套接口（`fetch` / `crawl` / `extract`）封装五种爬虫引擎，**一个环境变量切换**， > 上层叠加 **DeepSeek 结构化抽取**（网页 → JSON）。可从**代码库 / MCP 服务 / Claude Code SKILL** > 三种入口驱动 —— 三者都是同一个 `CrawlManager` 的薄封装。 `5 种后端` · `零依赖默认引擎` · `1 个变量切换` · `网页 → JSON` · `3 种入口` · `纯 CPU` · `实测可爬全球 Top10 的 9 个` > 上面这段演示由 [`demo.tape`](demo.tape) 用 [vhs](https://github.com/charmbracelet/vhs) 录制（`vhs demo.tape`）。

Tools (13)

AGENTCURL_LEARNCRAWL_BACKENDDEEPSEEK_API_KEYDEEPSEEK_MODELFIRECRAWL_API_KEYJINA_API_KEYRESPECT_ROBOTSROUTER_MODEbrowsercrawl4aifirecrawljinastatic

agentcurl

Quick Install

README Excerpt

Tools (13)

Related Browser/Web Servers