Turns URLs into clean markdown, runs web searches with full content extraction, or fact-checks claims by searching and citing sources. All requests route through Jina's infrastructure so your server IP stays hidden, which matters if you're scraping at scale or dealing with geo-restrictions. The grounding mode is interesting for RAG pipelines that need citation trails. Free tier gives you 10M tokens without signup. CSS selectors let you target specific page regions or strip out navigation cruft. The screenshot format and proxy options are nice touches. Pricing is straightforward at half a cent per page for standard extraction, though the ReaderLM model costs 3x more and grounding burns through 300K tokens per request.
npx skills add https://github.com/sundial-org/awesome-openclaw-skills --skill jina-reader