Installation
Install scrape-sdk, the CLI, and the MCP server.
Package
npm install scrape-sdk
# or
bun add scrape-sdkAgent skill
npx skills add SoulSniper-V2/scrape-sdk --skill scrape-sdkSubpath exports
| Subpath | What it is |
|---|---|
scrape-sdk | scrape, fromEnv, createScrapeClient, types, errors |
scrape-sdk/firecrawl | Firecrawl v2 / Keyless adapter |
scrape-sdk/tinyfish | TinyFish Fetch/Search + opt-in Agent adapter |
scrape-sdk/jina | Jina Reader + Search |
scrape-sdk/tavily | Tavily Extract + Search |
scrape-sdk/spider | Spider /scrape and /crawl |
scrape-sdk/browserbase | Browserbase Fetch API |
scrape-sdk/local | Cheerio + turndown |
scrape-sdk/ai | createTools → scrape_url, search_web, map_site, crawl_site, extract_json, optional run_web_agent |
MCP is a separate package: npx scrape-sdk-mcp.
Machine-readable docs
Do not scrape the HTML docs. Use:
/llms.txt— site index/docs/<path>.md— one page as Markdown/feeds/docs.jsonl— one JSON object per page
See For agents.