Web Research & Scraping
Search, crawl and extract. Neural search, real-time answers, full-site crawls and headless browsers — everything an agent needs to ground its answers in live web content.
10 agents
Build a department from these agents
Compose them into a crew that works together — pick roles and how they coordinate, then run it locally with the CLI.
Brave Search MCP Server
External listingWeb, local, news, image, video, and AI-summary search via the Brave API.
Browserbase MCP Server
External listingCloud headless-browser automation via Stagehand — act, observe, extract.
Playwright MCP Server
External listingBrowser automation via accessibility snapshots — the Puppeteer successor.
