I built Bookmark GPT Pro to solve my own problem: I had hundreds of bookmarks but Chrome's search only does keyword matching.
What it does: - Full-content search - searches titles, headings, descriptions, and full page text (not just titles) - AI chat interface to ask questions about your bookmark collection - Example: search "performance" and find "Making React Apps Faster" even if "performance" is only in the page content
Technical approach: - Weighted keyword search (title=50pts, headings=25pts, content=10pts) - AI chat powered by OpenAI or Ollama - Ollama support means you can run completely local/offline - No server-side storage - privacy-first design
Why this matters: Traditional folder organization breaks down at scale. Chrome only searches bookmark titles. Full-content search finds bookmarks based on what's actually on the page.
Privacy: With Ollama, everything runs on your machine. For cloud models, we only transmit encrypted queries - no bookmark storage.
Chrome Web Store: https://chromewebstore.google.com/detail/bookmark-gpt-pro/de...
Troubleshooting Guide: https://g-savitha.github.io/bookmarkgptpro-info/troubleshoot...
Built this over a few months of weekends. Would love HN's feedback on the approach and implementation!