firecrawl/cli

What it solves

Firecrawl CLI provides a command-line interface to search, scrape, and crawl the web, specifically optimized for AI agents and developers. It simplifies the process of turning unstructured web content into clean, AI-ready markdown or structured JSON, while also providing specialized indices for developer documentation and scientific research papers.

How it works

The CLI acts as a bridge to the Firecrawl API, allowing users to perform actions like scrape (extracting content from URLs), search (web search with optional scraping), and crawl (mapping and indexing sites). It includes a specialized developer command for searching GitHub issues, PRs, and READMEs, and a research command for searching a massive index of scientific abstracts (arXiv, PubMed, etc.). Additionally, it can install "skills" and MCP servers into AI coding agents (like Cursor or Claude Code) to make Firecrawl the default web provider for those agents.

Who it’s for

• AI Developers: Those building LLM applications that need high-quality web data.
• Coding Agents: Users of AI IDEs (Cursor, Windsurf) who want their agents to have better web search and scraping capabilities.
• Researchers: People needing to programmatically search and analyze scientific literature.

Highlights

• AI-Ready Output: Converts web pages to clean markdown or structured JSON using schemas.
• Agent Integration: Directly installs capabilities into AI coding agents (Claude Code, Cursor, etc.) via setup commands.
• Specialized Indices: Dedicated search for developer sources (GitHub/Docs) and a 43M+ research paper index.
• Flexible Scraping: Supports JS-rendering wait times, screenshotting, and PII redaction.
• MCP Support: Includes an MCP server for seamless integration with supported editors.