Top all-time · #9025
gpk blog
Personal blog · blog.gpkb.org · on HN as globular-toastprogrammingsystems
gpk blog is a personal blog at blog.gpkb.org about programming and systems. Since 2024, 9 different Hacker News readers have linked it 41 times, most often “Just Send Me the Prompt (2025)”. It is still writing: the last post was on 2026-10-08, and you can follow it by feed.
Too few sampled comments (3) to say how often HN recommends it.?
Why HN links it
HN readers link this blog for its take on AI-generated writing and engineering practice, such as the "Just send me the prompt" argument that AI-written articles add little beyond the prompt.
Summarised by Claude from a sample of 3 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
My ramblings about emacs, the internet and everything
Most-linked posts?
- Just Send Me the Prompt (2025)5 people · 15 mentions · first linked 2025-09-23
- Run AI Agents in Lightweight Sandboxes2 people · 10 mentions · first linked 2025-08-31
- Git Is a Version Control System2 people · 3 mentions · first linked 2024-03-11
- My TrueNAS Core (FreeBSD) Homelab2 people · 2 mentions · first linked 2026-03-08
What Hacker News says?
I invented a new technique that cuts down on the AI bill. I call it "just send me the prompt": https://blog.gpkb.org/posts/just-send-me-the-prompt/
Comment linking /posts/just-send-me-the-prompt/
on “You did this with an AI and you do not understand what you're doing here”Submitted Kubernetes – Ingress vs. Load Balancers as a story · 3 points
HN discussionYep, I wrote about it here: https://blog.gpkb.org/posts/just-send-me-the-prompt/
Comment linking /posts/just-send-me-the-prompt/
on “I don't want to read what you didn't write”Submitted Git Is a Version Control System as a story · 3 points
HN discussionI tried this but it's incredibly annoying as you'll get a mixture of file ownerships and permissions. Instead I run it in bubblewrap sandbox: https://blog.gpkb.org/posts/ai-agent-sandbox/
Comment linking /posts/ai-agent-sandbox/
on “Claude Code 2.0”Submitted Run AI Agents in Lightweight Sandboxes as a story · 2 points
HN discussionI'm a big fan of keeping tons of bash history. I have an idea for your tool: allow per-project history tracking. I do this myself for some projects using some bash history hacks: https://blog.gpkb.org/posts/project-local-bash-history/
Comment linking /posts/project-local-bash-history/
on “I quit my job to work full time on my open source project”Submitted Just Send Me the Prompt (2025) as a story · 1 points
HN discussionI used this for my homelab before I decided my life was too easy and switched to k8s+cert manager. I use traefik[0] to handle multiple docker-compose projects in development, but this is pretty much a drop in replacement and gets you HTTPS if you have a domain and your registrar has an API. I ran a single docker-compose file with everything in it behind a single reverse proxy for quite a while. [0] https://blog.gpkb…
Comment linking /posts/multiple-web-projects-traefik/
on “Caddy – The Ultimate Server with Automatic HTTPS”Submitted Just Send Me the Prompt as a story · 1 points
HN discussionThat's quite unnecessary. You could just mount "all tools and services" from your base system without the complicated Dockerfile: https://blog.gpkb.org/posts/ai-agent-sandbox/
Comment linking /posts/ai-agent-sandbox/
on “Zed is now available on Windows”Submitted My TrueNAS Core (FreeBSD) Homelab as a story · 1 points
HN discussionIndeed, you can use pmtiles files on a completely static site which is the main benefit for me. I wrote it up on my blog: https://blog.gpkb.org/posts/hugo-static-site-maps/ I have since started making my own pmtiles from pbf extracts using planetiler, though. I should add that as an addendum...
Comment linking /posts/hugo-static-site-maps/
on “2024: The year of the OpenStreetMap vector maps”Submitted Run AI Agents in Lightweight Sandboxes Using Bubblewrap as a story · 1 points
HN discussionMake is one of those things that I'm really glad I learnt at the beginning of my career. Not because I use it much any more, but because it showed me the power of a declarative system over an imperative one. I also realised at one point how naturally the idea extends to other tasks that I do. Going by the picture at the top of this site, it seems the author realised a similar thing to me: you can understand food rec…
Comment linking /posts/cooking-with-make/
on “Learn Makefiles”Whoa, thanks. I had my eye on Caddy for a local reverse proxy but can't be doing with maintaining another big config file. With the labels method the reverse proxy becomes "set and forget" and you're probably going to be editing bundled compose files to get rid of port conflicts anyway (or using override like https://blog.gpkb.org/posts/multiple-web-projects-traefik/ )
Comment linking /posts/multiple-web-projects-traefik/
on “Lcl.host: fast, easy HTTPS in your local dev environment”There are some more obvious things you should do (in order): 1. Have backups. You are running software all the time that can corrupt your files either maliciously or, more likely, accidentally. It doesn't really matter where it comes from, 2. Get into the habit of running things in sandboxes. You don't need anything magical here, a separate (unprivileged) user account is a good enough sandbox for many things. I outl…
Comment linking /posts/calibre-rootless-install/
on “Vet is a safety net for the curl | bash pattern”What about the fact that almost all websites that I want to search in hijack the quick search key now? https://blog.gpkb.org/posts/lamenting-firefox-quick-find/
Comment linking /posts/lamenting-firefox-quick-find/
on “A small change to improve browsers for keyboard navigation”It's kinda hilarious to watch people make themselves redundant. Like you're essentially saying "you don't need me, you could have just asked ChatGPT for a review". I wrote before about just sending me the prompt[0], but if your prompt is literally my code then I don't need you at all. [0] https://blog.gpkb.org/posts/just-send-me-the-prompt/
Comment linking /posts/just-send-me-the-prompt/
on “"ChatGPT said this" Is Lazy”So basically `scp -r` (recursive remote copy). It's what I call "dumb version control" and I've observed many people not going beyond this with git: https://blog.gpkb.org/posts/git-is-a-version-control-system/
Comment linking /posts/git-is-a-version-control-system/
on “A TUI Git client inspired by Magit”Sigh... Just Send Me the Prompt: https://blog.gpkb.org/posts/just-send-me-the-prompt/ If I needed this LLM output, which I probably don't, but let's say I do, I can just generate this myself. It's disrespectful to others to generate LLM output for them. Just send them the prompt. We all have access to these tools. It's like sending someone a pre-masticated apple. I have my own teeth, thank you.
Comment linking /posts/just-send-me-the-prompt/
on “Show HN: Build the habit of writing meaningful commit messages”I documented what I came up with for Claude Code: https://blog.gpkb.org/posts/ai-agent-sandbox/ However, I couldn't get this to work for Codex, it kept failing at the auth bit and I couldn't figure out how to fix it. Anyone got a working solution for Codex?
Comment linking /posts/ai-agent-sandbox/
on “Using bubblewrap to add sandboxing to NetBSD”I didn't know about using elisp in the replacement! Can't wait to need to use it. You can also do this replacement across an entire project and emacs also uses "smart" caps by default so that case is preserved from match to replacement. I wrote about it here: https://blog.gpkb.org/posts/emacs-regexp-replace/
Comment linking /posts/emacs-regexp-replace/
on “Elisp in Replacement String”I use this. I run it in a sandbox[0]. I run it inside Emacs vterm so it's really quick for me to jump back and forth between this and magit, which I use to review what it's done. I really should look into more "native" Emacs options as I find using vterm a bit of a clunky hack. But I'm just not that excited about this stuff right now. I use it because I'm lazy, that's all. Right now I'm actually getting into woodwor…
Comment linking /posts/ai-agent-sandbox/
on “OpenCode – The open source AI coding agent”I wrote about this a year ago: https://blog.gpkb.org/posts/just-send-me-the-prompt/ Sending LLM outputs to people is worse than useless, because you don't even know what they prompted with. For all you know it was "write me a paragraph justifying my opinion".
Comment linking /posts/just-send-me-the-prompt/
on “Don't be a meat proxy”> If you don't have anything to add, it's okay to say so. "No strong opinion here" is a real, helpful reply. This is really useful, I wish more people would understand this: you don't have to have an opinion on everything. In fact, you probably shouldn't have an opinion on most things. My own take on the AI issue: https://blog.gpkb.org/posts/just-send-me-the-prompt/
Comment linking /posts/just-send-me-the-prompt/
on “Don't Paste the AI, please”On Linux, you can skip Docker and use bubblewrap. Some inspiration: https://blog.gpkb.org/posts/ai-agent-sandbox/
Comment linking /posts/ai-agent-sandbox/
on “Docker Sandboxes – Disposable, isolated sandboxes for AI agents”In my experience, this already starts at a too high level. New devs these days don't even seem to come with a basic idea of what version control is. They think of git more as a workflow used to get code into production. This is why I wrote this article: https://blog.gpkb.org/posts/git-is-a-version-control-system/
Comment linking /posts/git-is-a-version-control-system/
on “The Guide to Git I Never Had”Related: https://blog.gpkb.org/posts/just-send-me-the-prompt/ (I could have sworn there was a popular HN submission a while back of this or a similar blog post, but damned if I can find it now.)
Comment linking /posts/just-send-me-the-prompt/
on “Don't Make Me Talk to Your Chatbot”I liked this article, and I see a lot of other commenters didn't, so I'll give my take: When starting on a new codebase, how do you make yourself into a helpful contributor as quickly as possible? I go straight for the humans and their human docs. What problem was the system originally built to solve? What was the original design, and what were its biggest problems? Who is currently using it? If you know these, read…
Comment linking /posts/just-send-me-the-prompt/
on “AI demands more engineering discipline. Not less”