Charitably, it could be an AI readme and hand rolled code, but it definitely is a smell.
Comment on I built a self-hosted period tracker because I couldn't find one worth using
rimu@piefed.social 2 days ago
I was going to recommend this to someone I know but when I realised your readme.md is entirely AI-generated, I guess the whole project is probably vibe-coded. I can’t in good conscience recommend someone trust their health data to a vide-coded app because they tend to have security problems.
Also all ai-generated code is public domain so your AGPL license is kinda empty. Might as well use MIT.
CameronDev@programming.dev 2 days ago
rimu@piefed.social 2 days ago
Yeah there are other signs too. Look at those commit messages, all vague, all perfectly capitalized. All with a nice long description with bullet points.
No one does that in a project they’re building for themselves.
EdTheMessenger@lemmy.world 2 days ago
Judging code quality by use of LLM in a documentation and commit messages is weird.
While I write all of my code myself and I’m against vibe coding etc., there is one place where I let a LLM write for me: readmes, commit messages and Javadoc comments.
I’m know how to write code but at the same time I’m shit in both my native language and even more so in English. So I let Language Models write natural language texts for me and just fix them when necessary. My documentation is more clear, grammatically correct and more detailed than in any of my previous projects, and I can focus on writing code.
And I wouldn’t say “No one does that in a project they’re building for themselves”. I do that for projects that only I will ever see, and OP shared his project with others, so it’s great that he included a clear documentation
terraincognita@lemmy.world 2 days ago
I answered earlier, that I use AI and this is just a commit skill for an agent.
helix@feddit.org 2 days ago
No one does that in a project they’re building for themselves.
Speak for yourself, I always did that and I found it easier with LLMs nowadays.
I hate most AI shite with a passion but when it helps my colleagues write commits which are more than “add stuff”, “fix some things” I’m fine with it.
I rarely use AI to generate code, usually only when I need a starting point. It’s much easier to unfuck AI code than to stare blankly at a screen for an hour. I’d never commit code I don’t fully understand or have read to the last byte.
I hope OP is doing the same. LLMs fail at 90% of coding tasks for me but for the other 10% (mostly writing tests, readmes, boilerplate) it’s really OK for productivity.
Ethics of LLMs aside, if you use them for exactly what they’re built for – being a supercharged glorified autocomplete – they’re cool. As soon as you try to use them for something else like “autocompletion from zero” aka “creativity”, they fail spectacularly.
mortalic@lemmy.world 2 days ago
Thanks for doing this, I was debating doing the same. It needs to exist.
rimu@piefed.social 2 days ago
F-Droid has Drip, Bluemoon and Periodical.
terraincognita@lemmy.world 2 days ago
Yes, I’m aware of those apps. They’re great local-first mobile trackers. Ovumcy explores a slightly different approach - a self-hosted web app that can run on infrastructure you control and be accessed from multiple devices.
terraincognita@lemmy.world 2 days ago
I do use AI tools while developing this project, but I also have a BSc in Computer Science. AI is a productivity tool.
Security is something I take seriously, especially since the project deals with health data. All code has test and you’re welcome to inspect the repository yourself or point out any specific security concerns if you notice them.
Regarding licensing: the AGPL license applies to the project as a whole regardless of the tools used to write parts of the code.
If you have concrete technical feedback or security issues, I’d genuinely appreciate it.
sonofearth@lemmy.world 2 days ago
You should add a disclaimer stating that you have used an LLM. I have done so for a tool I built with an LLM that I needed, because I don’t know jackshit about coding and I am not gonna pretend I do.
terraincognita@lemmy.world 2 days ago
Partially agree, but I do know how to code and use it as a tool.
chicken@lemmy.dbzer0.com 2 days ago
But if OP does know and apply that knowledge to what they are doing, it’s not the same thing and doesn’t make sense to have the same disclaimer.
Serinus@lemmy.world 2 days ago
It’s not realistic to expect no AI assistance in coding in 2026.
It’s also not a stand-in for a human. There’s a huge field of gray where it’s unclear how much of it was fully vibe coded vs how much is carefully hand reviewed and/or written.
I’ve been a professional developer for decades and I’ve done both. Obviously I’ve hand coded stuff for many years. The fully vibe coded stuff is personal, to test and learn the capabilities of the tech. My professional stuff I watch much more closely, and I’m much more targeted in what I’m having the AI do.
That said, if I were gonna use this I’d actually review the code. I’m not recommending this guy’s stuff, but you can’t rule it out on the basis of ai assistance alone.
CameronDev@programming.dev 2 days ago
It may not be a stand in for a human, but that’s exactly how many of these vibe coded projects are. It’s not unreasonable to ask the developer to spend 30 seconds to describe how they use these tools.
teawrecks@sopuli.xyz 2 days ago
A bunch of people who couldn’t tell their left shift from their right shoelace think you don’t know what you’re talking about lol.
I agree, to a person who knows the machine, an AI is like a compiler: you know the output you’re going for, the tool helps you get there faster. Expecting you to do something the slow way because someone else doesn’t know how to code is nonsense. There is a massive difference between using it as a tool, and blindly taking generated code.
If the internet existed in the 70s, I bet people would have asked for a disclaimer on compiled assembly.
pimento64@sopuli.xyz 1 day ago
Pathetic. You have a serious skill issue, that much is true, but you need to keep it to yourself instead of sharting it out into society. Code better or go get a job you’re more qualified for, like operating a tollbooth.
iamthetot@piefed.ca 2 days ago
Guess I’ll stick to unrealistic software then.
terraincognita@lemmy.world 2 days ago
You can see that I use some of metrics, like test coverage, estimates and so on to prove its validation as potentially serious project, that will grow from a pet one.
Tibi@discuss.tchncs.de 2 days ago
Testcoverage by ai generated Tests is close to worthless. “Tests are only as good as the person writing them”
Did you generate your tests?
Zak@lemmy.world 2 days ago
Why?
It makes sense to try to give users an idea of how robust a project is, but the exact details of the tools involved in its creation rarely add much to that. It gets a little weird with LLMs because they allow someone with no programming skill to create software that appears to work, which ought to be disclosed; “I don’t know what I’m doing and I asked a robot to make this” does indicate unreliable code. A skilled developer having an LLM fill in some extra test cases, on the other hand can only make the project more robust.
IanTwenty@piefed.social 2 days ago
The danger being raised with the licensing is that you can’t license something if you’re not considered to be the author. There are growing examples of courts and lawmakers determining AI output to be public domain:
https://www.theregister.com/2026/03/06/ai_kills_software_licensing/
This is an evolving, global situation and hard to know what to do right now. I think what you’ve got is fine though - you’ve made it clear your intention is to license with AGPL. It’s just that depending on the jurisdiction it might be public domain instead.
This is another reason to be clear about the use of AI in the README so your users can make an informed decision.
terraincognita@lemmy.world 2 days ago
I agree, though there is a difference in case you rovided and mine. It is a human-directed work. Thousands of libraries, Kubernetes, Kubernetes still live and license is valid.
militaryintelligence@lemmy.world 2 days ago
How does AI help with productivity? I’ve gotten so many false answers that I quit trusting it
prenatal_confusion@feddit.org 2 days ago
Imagine you are on the ground under your car and need a different tool. You ask for it and somebody hands it to you. That person is young and inexperienced. It is up to You to check if it’s the right tool, and if not pass it back (and in this example tell the person about the error and help them correct it).
And sure, You can always crawl out and get the tool yourself and sometimes that is the only option and in coding terms in my opinion best practice. But you can be faster with your helper. Use it appropriately and see how it affects your work. And that’s the point, your work. Don’t pass responsibility or thought off to AI.
militaryintelligence@lemmy.world 1 day ago
Ok but give me a wrench example in coding terms. What wrench are you wanting? I really am curious
moriquende@lemmy.world 2 days ago
Because it’s able to write boilerplate faster than a human. And because it’s able to perform refactorings that are not possible with IDEs or regex due to their lack of structure. Also because you can ask it to review your files and it does find bugs that would otherwise be missed at first. There’s a huge difference between vibe-coded slop and using the tools available to you effectively.
SorryQuick@lemmy.ca 2 days ago
I use it all the time as a kind of brainstorming tool.
“I want to do X (and details), can you tell me what tools or algorithms are available to me? List their pros and cons and give me some comparison”
Or on a somewhat recent project of mine, I has to effectively stub an entire library (but didn’t have to be done well), so I just told AI “take this page (the docs) and generate empty function stubs from it”. It doesn’t need to be high quality since it doesn’t run, it’s just to fool the dotnet engine.
I also tend to ask AI to add logging to some functions, since it’s annoying to do and impossible to mess up.
militaryintelligence@lemmy.world 1 day ago
Best answer I’ve read. I get why it’s used, I just don’t trust it because there’s got to be a catch as hard as companies are pushing it. It’s available for free, so we must be the product somehow
jamoquanty@mastodon.social 2 days ago
@militaryintelligence @terraincognita
They are tools like any other, it's how you use them that matters.
xvapx@lemmy.world 2 days ago
No, they’re tools different from any other due to their nondeterministic nature. That, coupled with hallucinations, are the big differences that make me automatically dismiss anything using these tools.
dogs0n@sh.itjust.works 2 days ago
I’m guessing you let the AI make the tests and everything, which wouldn’t give me much reassurance that any of the code is good. Sadly AI will jump through any hoops it can to get tests to pass if it can’t get the code working.
I think people who let AI run wild to create a whole app should write the tests themselves or at least only with line completion (jusdging by a quick look at the project files, I am guessing an AI did everything).
Could be food for thought?