Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. An agent is not an extension of the employee who created it(getlago.com)
    discuss
  2. Agent Communication Protocol(agentcommunicationprotocol.dev)
    discuss
  3. Search is a 2.9 mb browser for the Mac based on WebKit(officecommun.com)
    discuss
  4. Tokonomics–AST context compiler to stop hitting Claude/GPT rate limits in VSCode(visualstudio.com)
    discuss
  5. First Principles Thinking(sunilsadasivan.com)
    discuss
  6. EA Safety(venkateshrao.com)
    discuss
  7. Why 'What's Opera, Doc?' Looks Like That(animationobsessive.substack.com)
    discuss
  8. Why do common Rust packages depend on C code?(langdev.stackexchange.com)
    discuss
  9. Show HN: Photos by Strangers (reviving an old, self-coded project)(serendi.pt)
    discuss
  10. After multiple deaths and injuries, NHTSA is investigating comma.ai(arstechnica.com)
    1comments
  11. I Built AI Clones of My Coworkers. Things Got Weird(wired.com)
    discuss
  12. IonStack: 1-Click Browser to Kernel Full-Chain to Get Root Shell on Android 17(blackhat.com)
    discuss
  13. Energy Sharing and the Grid(netbound.io)
    1comments
  14. Call an API from Every Row in DuckDB, Without the Loop(query.farm)
    discuss
  15. Rails World 2026 Opening Keynote [video](youtube.com)
    1comments
  16. Global DRAM Module Revenue Jumped 59% to $21.2B in 2025(guru3d.com)
    discuss
  17. Ten years of Postgres logical replication(tapoueh.org)
    discuss
  18. You Can See Everything – Official Trailer [video](youtube.com)
    discuss
  19. What to Know About JavaScript in 2026(master.dev)
    discuss
  20. Lightspeed: Deterministic agent harness for Temporal (in Rust)(github.com/smartcomputer-ai)
    discuss
  21. Confronting Unknowns: A Framework for Anomalous Aerospace Events(sensemaking.wtf)
    discuss
  22. Proxy and Reflect(piccalil.li)
    discuss
  23. AMD-AGI/HyperLoom – Agentic system that auto-optimizes LLM workloads on AMD GPUs(github.com/amd-agi)
    discuss
  24. Gemini 3.8 text-to-speech says hello(blog.google)
    discuss
  25. I See Dead Cars: A Mailman's Daily Trip Down Memory Lane(caranddriver.com)
    discuss
  26. Paste.sh – Encrypted Pastebin(paste.sh)
    discuss
  27. Why Human Informants Still Matter in the Age of AI Surveillance(smallwarsjournal.com)
    discuss
  28. You Can See Everything – Official Trailer – A24 [video](youtube.com)
    discuss
  29. Valve's Linux Steam client now runs on an Android phone. No root(the412banner.github.io)
    discuss
  30. Funding MIT Software with Commercial-Use Binaries(yaak.app)
    discuss

Show HN: Attention – A reading extension that uses your AI-generated profile

1 pointsby 1h agochromewebstore.google.com
1 comments
1h agoHN ↗

First of all: I’m not a programmer and this project is vibe-coded 100%. In recent months, I have been experimenting with personal knowledge management (PKM) and reading many articles. I realized how much time I spend reading articles about topics I already know pretty well. These articles may be good but there is nothing really new in them for me. On the other hand, we say so much to AI assistants about our interests, goals, tasks, knowledge, and questions. So why not to leverage the context to choose what articles should be read, what skipped and where useful passages can be found? Attention is a Chrome extension I developed. First, you request your AI assistant to draft your profile using available context, correct it and import into the extension. Then, Attention leverages the profile and suggests what to read, what to skip and what passages may be useful for you. There are currently two modes: heuristic analysis that takes place locally, and AI analysis which requires API key. I will be very happy to hear your thoughts on it. P.S. Profile creation itself gave me a feeling of “taking back what is mine from the platform” :))