Hacker News: natloz

New comment by natloz in "Show HN: Agent that refuses to run commands without human approval"

natloz — Fri, 01 May 2026 14:23:45 +0000

I feel this is not the trajectory we want to go with automation. Perhaps better checks and balances within the automation, or "thresholds" that trip breakers would be a good approach?

New comment by natloz in "Show HN: I wrote a DOOM clone in my own programming language"

natloz — Thu, 30 Apr 2026 04:07:12 +0000

Got to love these classic games that were so amazing in their glory days and still fun today!

New comment by natloz in "Show HN: CrabPDF – privacy-first PDF editor that edits real text"

natloz — Mon, 27 Apr 2026 16:02:09 +0000

Interesting approach. I think even if you assure that it is not in the cloud, most people will associate uploading something to the web-browser as being in the cloud, may be a tough hurdle. Cool idea however, nice work!

New comment by natloz in "I built an agent that breaks your AI agents before someone else does"

natloz — Sun, 26 Apr 2026 21:33:34 +0000

Cool, sounds like an interesting experiment!

New comment by natloz in "I built an agent that breaks your AI agents before someone else does"

natloz — Sun, 26 Apr 2026 18:45:36 +0000

What happens if you point Nyx at itself, who breaks first!

New comment by natloz in "Show HN: Good AI Task – a tool for asking AI what it can and can't do"

natloz — Sun, 26 Apr 2026 18:41:17 +0000

Does it account for its ability to think it can do what it actually can't (hallucinations, biased, etc...)

New comment by natloz in "Show HN: Built a daily game where you sort historical events chronologically"

natloz — Fri, 24 Apr 2026 19:18:12 +0000

I saw images (I think?), nothing looked off!

New comment by natloz in "Show HN: Built a daily game where you sort historical events chronologically"

natloz — Fri, 24 Apr 2026 04:34:20 +0000

Wow, just played my first game and I wild-guessed it and got it right in 1 attempt. Maybe time to go buy a lotto ticket...I honestly had no clue! Good job on the easy-to-use interface.

New comment by natloz in "Show HN: Gemini Plugin for Claude Code"

natloz — Thu, 23 Apr 2026 19:27:43 +0000

I definitely find value when working with models to run things by for another perspective. My favourite moment was when I ran a plan that Replit made by Claude, Claude had some opinions about it, brought them back to Replit, and Replit's response was "Claude is over-reacting", made me chuckle. I felt like such a middle-man.

New comment by natloz in "Show HN: A free tool for non-technical folks to easily publish a website"

natloz — Thu, 23 Apr 2026 18:09:51 +0000

As I read your post I see terms like GitHub, OAuth, OAuth flow; these are already technical terms that many people would not understand. How could you describe this to your target audience? Just curious.

New comment by natloz in "Show HN: A lightweight way to make agents talk without paying for API usage"

natloz — Wed, 22 Apr 2026 15:36:01 +0000

Any methods to reduce the cost of using AI (which can be exorbitant) is welcome!

New comment by natloz in "Show HN: Humanoid.js – One HTML file that scores how human your clicks look"

natloz — Wed, 22 Apr 2026 14:21:29 +0000

Is this to stave off bots?

New comment by natloz in "Show HN: Self-hosted Raspberry Pi wall display (no cloud, no subscription)"

natloz — Tue, 21 Apr 2026 20:53:08 +0000

The Raspberry Pi is so versatile once you get creative with it, nice work!