Summarize Hacker News Comments with AI Browser Plugin
Command Palette
Search for a command to run...
Comments
Join the discussionNo comments yet. Be the first to comment.
Search for a command to run...
No comments yet. Be the first to comment.
I have been working with Ann Catherine on a Hacker News specific browser extension and I would like to share the details here.
We both love the rich discussions in HN, but navigating long posts with multiple threads can sometimes be challenging. You get lost in the depth and breadth of the discussion. There are a few AI-based summarization tools, but they typically focus on the article rather than the comments, and thus missing the essence of threaded discussions.
So, we built Hacker News Companion, a browser extension that makes it easy to navigate discussions and understand diverse viewpoints using AI tools of your choice. The goal is simple - make the comments the true hero in HN.
Summarize any thread or full post
See the key points, takeaways and notable quotes in summary panel
Jump to specific threads from the summary with auto-generated back-links - allowing you to re-engage in the discussion
Bring your own AI models - local or cloud-based:
Local: Ollama with any local model (privacy-focused, runs locally)
Cloud: OpenAI/Claude - use your API key
Browser: Chrome's built-in AI powered by Gemini Nano
Vim-inspired keyboard navigation
home row keys to move through posts and comments (home pages and comments page)
j/k for next/previous, h/l for parent/child comments
o to open posts, c for comments
? or / for all shortcuts, s to toggle the summary panel
Navigate author comments: use [ and ] keys or arrow buttons next to username
Hover usernames to see karma and profile info
See demo video for more details:
You can try it immediately by installing from Chrome Web Store.
No signup needed. The extension preserves privacy (no analytics/tracking) and the code is open source - https://github.com/levelup-apps/hn-enhancer.
Technical details - our learnings about summarizing threaded discussions and comparing the performance of various models are detailed in the project wiki.
Try it out and let us know how it works for you. Would love your feedback and contributions!
Love to hear your feedback here or in the HN thread.
