Web scraping
Firecrawl-backed page extraction. Markdown or a link list, standard or enhanced mode, billed in credits with no Firecrawl account of your own.
Your agents call them. Almost all of them need no account and no API key at all; the one that acts as you connects once with OAuth, and the credential never reaches the agent.
Firecrawl-backed page extraction. Markdown or a link list, standard or enhanced mode, billed in credits with no Firecrawl account of your own.
Perplexity-backed public-web search with recency and domain filters and a result limit. No Perplexity account of your own, billed in credits per call.
Find professionals by role, employer, skill, or location. Up to 20 source-backed profiles per call from Perplexity, billed in credits, with no account of your own.
Profiles, tweets, timelines, mentions, and recent search through the X API. One OAuth connect, no X developer account, no key in the agent.
Google Maps geocoding, directions, and place search plus OpenStreetMap data and rendering. No Maps project, no key, billed in credits.
Google Weather and Air Quality data: current conditions, hourly and daily forecasts, recent history. No account, and calls currently charge 0 credits.
Yahoo Finance data through a managed provider: search, profiles, quotes, and chart history. One credit per successful call, no brokerage or data account.
A managed browser session your agent controls and you can watch or take over from a live link. Survives the run, reclaimed after ten idle minutes.
Publish a static build directory to a public URL, with immutable version history and single-page-app routing. No hosting account, no deploy pipeline.