I'm a long-time reader of HN comment threads as an unusually effective way to get a lot of wisdom quickly about a wide array of topics that interest me. I built this tool to make that process even more efficient. It pulls the top stories, filtered for the ones with the most comments, and then lets you run an analysis of the comment thread. When you click "Analyze", it fetches full text of the entire comment thread, uses LLM to summarize key points, areas of consensus/disagreement, sentiment, etc (you choose short/medium/detailed length), and, finally, suggests context-relevant follow-up questions you can trigger.

I've been using it daily on my commute for a few weeks now as another/different way of consuming HN.

I’d love feedback — both on the idea itself and the execution. No signup is required to try it out for free. I used Lovable to build it, which was a lot of fun and a bit addictive for a little while.