Model Context Protocol

Cursor, Claude Code, and Codex already have WebSearch and WebFetch. Claude Code's WebFetch summarizes the page. Codex search defaults to a cache. This server is for the full markdown body, site map, crawl, and JSON extract.

{
  "mcpServers": {
    "scrape-sdk": {
      "command": "npx",
      "args": ["-y", "scrape-sdk-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "fc-...",
        "TAVILY_API_KEY": "tvly-...",
        "JINA_API_KEY": "jina_..."
      }
    }
  }
}

A copy lives at examples/mcp.json. The skill at skills/scrape-sdk/ ships a sibling mcp.json.

Tools:

• scrape_url — always. Full page markdown, default 20_000 chars, truncated + charCount. Named so it does not collide with host WebFetch.
• search_web — only if TAVILY_API_KEY or FIRECRAWL_API_KEY is set. Otherwise use the host WebSearch.
• map_site / crawl_site / extract_json — when a configured provider supports them. Hosts do not ship these.

fromEnv() reads FIRECRAWL_API_KEY (or FIRECRAWL_KEY), TAVILY_API_KEY, JINA_API_KEY, SPIDER_API_KEY, BROWSERBASE_API_KEY.