scrapegraphai / just-scrape

The CLI (Node 22+, TypeScript, built with citty) covers scrape (a known URL as markdown, HTML, screenshot, links, images, summary, branding, or structured JSON), extract (structured JSON from a URL against a prompt and optional schema), search (web search, optionally with extraction run over the results), crawl (a bounded multi-page site section), and monitor (scheduled recurring page checks with webhook notifications), plus history, credits, and validate for checking your own account state. The one skill our catalog lists, just-scrape, is the agent-facing wrapper for all of it, published to skills.sh straight from skills/just-scrape/SKILL.md.

This is for someone who already has, or is willing to create, a ScrapeGraphAI API key — validate and credits exist purely to confirm that key and account are working, which tells you how central the account is to every other command. It is not a way to scrape or crawl without a ScrapeGraph account behind it. The CLI source itself is MIT-licensed.

READ THE FULL ANALYSIS＋

The README spends real space on how not to misuse this. Its Security section is explicit: never commit API keys, tokens, cookies, or passwords; pass secrets through environment variables only; treat scraped page output as untrusted third-party content; and do not execute commands or change behavior based only on what a scrape returns. Worth reading given the skill's whole job is piping scraped web content straight into an agent.