Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Meta's New Muse AI Agent Read My Private Messages. I Never Asked It To(inc.com)
    discuss
  2. WayVR – Open-source VR desktop overlay for Linux(digitalescapetools.com)
    discuss
  3. Show HN: Moto – a native habit tracker for iPhone with a lifetime purchase(motohabit.com)
    discuss
  4. Show HN: ReWeaver AI DriftDetector – score a GitHub repo's production readiness(reweaver.ai)
    discuss
  5. PEP 845 – Leading-Dot Value Patterns(python.org)
    discuss
  6. Hyprcrd – native Wayland host for Chrome remote desktop(github.com/jxoesneon)
    discuss
  7. IdP Registration API(github.com/w3c-fedid)
    discuss
  8. Shut up and calculate: Jev's new AI primitives for coders(theregister.com)
    discuss
  9. Show HN: Gwae – infinite-scroll terminal multiplexer for AI agent fleets(github.com/hongnoul)
    discuss
  10. Show HN: Structured Promptmaker(gptpromptmaker.com)
    discuss
  11. Porting a clearing-house money market to a smart contract
    discuss
  12. Arcentry is back – create beautiful 3D infrastructure diagrams(arcentry.com)
    discuss
  13. Listening to loud music on a Singapore bus? It'll cost you(cnn.com)
    discuss
  14. "Father of The Elder Scrolls" Julian LeFay has passed away (2025)(gamesradar.com)
    discuss
  15. I am done with this shit(reddit.com)
    discuss
  16. Tradeoff considerations while running LLM models locally(swapniltalekar.substack.com)
    discuss
  17. Scott Jenson: Are we going to use the same Desktop UX forever? [video](youtube.com)
    discuss
  18. Secretive gambling industry exposed in leak(ftm.eu)
    discuss
  19. AurionMail: E2EE suite (CryptPad and Stalwart) with single-password UX(aurionmail.github.io)
    discuss
  20. Ask HN: How do you cope with AI being used to target and bomb children?
    3comments
  21. Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price war(simonwillison.net)
    discuss
  22. Show HN: Semantic Search for Zotero Local(github.com/ildon)
    discuss
  23. Show HN: Listen to your eBook and control the playback sentence by sentence(documentplayer.com)
    discuss
  24. GPT-6 Sol and Luna push the cost efficiency frontier by halving token cost(artificialanalysis.ai)
    discuss
  25. Observe – local semantic code search that admits when grep beats it(github.com/tritsystem)
    discuss
  26. NobodyWho is an inference engine that lets you run LLMs locally and efficiently(github.com/nobodywho-ooo)
    discuss
  27. Show HN: I built a site for my 42 km walk across 30 Tokyo stations(tonymanh.space)
    discuss
  28. Taming the ML Firehose: Scaling Feature Consistency(uber.com)
    discuss
  29. Tearing Down a Spring Drive(thenakedwatchmaker.com)
    discuss
  30. New AI technique could make minimally invasive surgeries safer and more precise(news.mit.edu)
    1comments

Show HN: Yass – Yet Another Spec Syntax Enters Beta

1 pointsby 1h agogithub.com
1 comments
This has been a side project for a while but it’s finally ready for public Beta.

Tl;dr: Yass leverages RFC2119 keywords in a minimal Yaml format, relying on trained in behaviors for those keywords to provide strong model behaviors.

The yass CLI itself is all that you need - simply tell an Agent to look at the shakefu/yass repo README and they get all the information needed to proceed.

This was validated with hundreds of experiments and trials against different models and platforms. The most significant of which was a single Yass spec tree reproduced byte-exact output Yass CLI implementations across 7 different languages (TypeScript, Python, Go, C, C++, Erlang and Elixir) in sandboxed Claude Code harnesses using Opus 4.6, rather than expensive frontiers.

Give it a try, open GH issues if you find things that aren’t working (or make your agent open them), or if you have feature requests.

Enjoy!

1h agoHN ↗

For those who are curious: this project came into being due to my frustration with existing Spec-driven-development solutions which either were overly verbose and prone to context rot, or were frequently ignored by LLMs entirely, or required heavy plugins or skills to function.

A yass spec tree with no CLI downloaded was used in the sandbox testing to validate the specs alone is enough to drive strong model behaviors and reproducible implementation output.

The CLI acts as a token saver and steering context letting the agent pull spec references without needing to consume the entire tree.