Credentials. The CRAWL4AI_KEY bearer token is billed for hosted requests.
Who it is forA developer whose agent needs to read web pages during research.
The problemAgents' built-in fetch returns messy HTML.
The outcomeScrape, extract and search tools available to the agent.
What you need first
- Python with pip
- Playwright Chromium (installed by crawl4ai-setup)
Steps
- Get a Crawl4AI key
- Run
claude mcp add --transport http crawl4ai https://api.crawl4ai.com/mcpwith the Authorization header - Ask the agent to scrape a page
- Use the Markdown in your task
Prompt for your agent
With the crawl4ai MCP server connected, scrape <url> to Markdown and summarize it. Confirm the summary only uses content from the page.
Caveat. The hosted API is optional; the open-source crawler runs free on your machine.
Use this in your AI agent
Ask your agent
Using the Deduped MCP, get the setup steps for unclecode/crawl4ai and "Give Claude Code a web scraping tool over MCP"
Add Deduped to Claude Code once
claude mcp add --transport http deduped https://dedupedai.com/mcp
Every Monday: the 5 hottest repos and exactly how to use them. Free, one email a week, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer