网站爬取
cf-crawl
by bill492
Crawl websites using Cloudflare Browser Rendering /crawl API. Async multi-page crawl with markdown/HTML/JSON output, link following, pattern filtering, and AI-powered structured data extraction. Use when crawling entire sites or multiple pages, building knowledge bases, extracting structured data from websites, or when web_fetch is insufficient (JS rendering, multi-page, authenticated crawls).
安装
claude skill add --url github.com/openclaw/skills/tree/main/skills/bill492/cf-crawl文档
Cloudflare /crawl
Async site crawler via CF Browser Rendering API. Start a job → poll for results → get markdown/HTML/JSON per page.
Quick Start
# Crawl a site (5 pages, markdown, no JS rendering = fast + free)
bash ~/clawd/skills/cf-crawl/scripts/crawl.sh "https://example.com" --limit 5 --format markdown
# With JS rendering (for SPAs, dynamic content)
bash ~/clawd/skills/cf-crawl/scripts/crawl.sh "https://example.com" --render --limit 10
# Start only (get job ID, poll later)
bash ~/clawd/skills/cf-crawl/scripts/crawl.sh "https://example.com" --limit 100 --start-only
# Poll existing job
bash ~/clawd/skills/cf-crawl/scripts/poll.sh <job-id>
Credentials
Stored at ~/.clawdbot/secrets/cloudflare-crawl.env:
CF_ACCOUNT_ID=<account_id>
CF_CRAWL_API_TOKEN=<token_with_read_and_edit>
Key Options
| Option | Description |
|---|---|
--limit N | Max pages (default 10) |
--depth N | Max link depth (default 10) |
--format markdown|html|json | Output format (default markdown) |
--render | Enable headless browser (default: off = fast fetch, free during beta) |
--include PAT | Wildcard URL pattern to include (repeatable) |
--exclude PAT | Wildcard URL pattern to exclude (repeatable) |
--external | Follow external domain links |
--subdomains | Follow subdomain links |
--source all|sitemaps|links | URL discovery method |
--json-prompt "..." | AI extraction prompt (with --format json) |
--json-schema file.json | JSON schema for structured extraction |
--timeout SEC | Max poll wait (default 300s) |
--output FILE | Write full results to file |
--raw | Output raw API response |
--start-only | Print job ID without polling |
Common Patterns
Crawl docs site for knowledge base
bash ~/clawd/skills/cf-crawl/scripts/crawl.sh "https://docs.example.com/" \
--limit 50 --depth 3 --format markdown --output docs.json
Crawl with URL filtering
bash ~/clawd/skills/cf-crawl/scripts/crawl.sh "https://example.com/" \
--include "/docs/**" --exclude "/docs/archive/**" --limit 20
AI-powered structured extraction
bash ~/clawd/skills/cf-crawl/scripts/crawl.sh "https://example.com/products" \
--format json --render \
--json-prompt "Extract product name, price, and description" \
--json-schema schema.json
Long-running crawl (background)
JOB_ID=$(bash ~/clawd/skills/cf-crawl/scripts/crawl.sh "https://big-site.com" \
--limit 1000 --start-only)
# Check later:
bash ~/clawd/skills/cf-crawl/scripts/poll.sh "$JOB_ID"
Cost Notes
render: false(default) — fast HTML fetch, free during betarender: true— uses Browser Rendering minutes (paid)format json— uses Workers AI tokens for extraction (paid)- Results cached in R2 with
--max-age(default 24hr)
API Details
See references/api-reference.md for full parameter documentation, response schema, and lifecycle details.
相关 Skills
前端设计
by anthropics
面向组件、页面、海报和 Web 应用开发,按鲜明视觉方向生成可直接落地的前端代码与高质感 UI,适合做 landing page、Dashboard 或美化现有界面,避开千篇一律的 AI 审美。
✎ 想把页面做得既能上线又有设计感,就用前端设计:组件到整站都能产出,难得的是能避开千篇一律的 AI 味。
网页应用测试
by anthropics
用 Playwright 为本地 Web 应用编写自动化测试,支持启动开发服务器、校验前端交互、排查 UI 异常、抓取截图与浏览器日志,适合调试动态页面和回归验证。
✎ 借助 Playwright 一站式验证本地 Web 应用前端功能,调 UI 时还能同步查看日志和截图,定位问题更快。
网页构建器
by anthropics
面向复杂 claude.ai HTML artifact 开发,快速初始化 React + Tailwind CSS + shadcn/ui 项目并打包为单文件 HTML,适合需要状态管理、路由或多组件交互的页面。
✎ 在 claude.ai 里做复杂网页 Artifact 很省心,多组件、状态和路由都能顺手搭起来,React、Tailwind 与 shadcn/ui 组合效率高、成品也更精致。
相关 MCP 服务
GitHub
编辑精选by GitHub
GitHub 是 MCP 官方参考服务器,让 Claude 直接读写你的代码仓库和 Issues。
✎ 这个参考服务器解决了开发者想让 AI 安全访问 GitHub 数据的问题,适合需要自动化代码审查或 Issue 管理的团队。但注意它只是参考实现,生产环境得自己加固安全。
Context7 文档查询
编辑精选by Context7
Context7 是实时拉取最新文档和代码示例的智能助手,让你告别过时资料。
✎ 它能解决开发者查找文档时信息滞后的问题,特别适合快速上手新库或跟进更新。不过,依赖外部源可能导致偶尔的数据延迟,建议结合官方文档使用。
by tldraw
tldraw 是让 AI 助手直接在无限画布上绘图和协作的 MCP 服务器。
✎ 这解决了 AI 只能输出文本、无法视觉化协作的痛点——想象让 Claude 帮你画流程图或白板讨论。最适合需要快速原型设计或头脑风暴的开发者。不过,目前它只是个基础连接器,你得自己搭建画布应用才能发挥全部潜力。