I built a site that finds book recommendations from tweets using Grok's X Search API
I wanted book recommendations from people I follow, but most sites just list books without any source.
So I built a site that searches public tweets for book recommendations using Grok's x_search API.
How it works:
- Uses Grok API with the x_search tool to search X/Twitter
- Extracts book mentions from tweets containing "great read", "must read", "favorite book", etc.
- Each recommendation links back to the original tweet as a source
Tech stack: Astro, Python extraction script, Grok 4 API with x_search, Cloudflare Pages
Currently tracking 150+ personalities with 1000+ book recommendations, all sourced from real tweets.
Would love feedback on the UX or suggestions on how to improve!
Trending on Indie Hackers
Great idea I wish you success
Thank you !
How did you buid the app ?
Specifically, which tools are you using ?
I'm using X API for the data extraction from tweets as Grok / X AI has access to them, Astro for the front-end
Nice! Using the x_search API as a data source is clever — it eliminates the rate-limit and structure-change headaches you'd hit with traditional scrapers. The 'extract from unstructured text' step is where most of these projects fail. Are you using Grok's LLM for the extraction, or custom parsing logic?
I am using a mix of both custom logic and Grok prompt for the extraction