Firecrawl is the context API for AI agents. This Cursor plugin lets your agent search, scrape, and interact with the web through the hosted Firecrawl MCP server, with the Firecrawl CLI as a fallback. It also crawls and maps sites, monitors pages for changes, and runs Alexandria data providers for structured records.
mcp.json- The hosted Firecrawl MCP server athttps://mcp.firecrawl.dev/v2/mcp-oauth. The first Firecrawl tool call opens a browser sign-in; no API key to paste.skills/- 13 skills, synced automatically from firecrawl/cli:firecrawl,firecrawl-agent,firecrawl-alexandria,firecrawl-crawl,firecrawl-developer-index,firecrawl-download,firecrawl-interact,firecrawl-map,firecrawl-monitor,firecrawl-parse,firecrawl-research-index,firecrawl-scrape,firecrawl-searchrules/install.mdc- CLI installation and authentication, for when the MCP server is unavailable
Install the plugin and approve the Firecrawl sign-in when Cursor asks (or open Cursor Settings > MCP). That covers every remote operation.
The CLI is optional, for when the MCP server is unavailable or for local files:
npm install -g firecrawl-cli
firecrawl login --browser- Search - Web, image, and news search with optional full-page scraping of results, plus matching Alexandria tools
- Alexandria - Discover catalogued data providers (official APIs, licensed publishers, Firecrawl indexes), read their contracts, and execute them for structured records instead of scraping pages
- Scrape - Single page content extraction to clean markdown, with JS rendering support
- Map - Discover all URLs on a site, with search filtering
- Crawl - Bulk content extraction from entire site sections
- Interact - Remote browser sessions for interactive pages (pagination, forms, logins)
- Agent - AI-powered autonomous web research and structured data extraction
- Developer and research indexes - Issues, merged pull requests, READMEs, docs, and published papers
- Monitor, download, parse - Watch pages for changes, save a site locally, parse local files