You don't. If you're using skills to force the AI to fullfill some must criterias, it's not going to work. Must criterias need deterministic checks -> be it hooks or what not.

This is also my biggest gripe with AI. I.e. for specifications, no matter what hype machine I tried, it never fulfilled my criterias, which are: easily verifiable, concise, small specs. Hence I built https://github.com/RicardoMonteiroSimoes/Yamlet initially for claude code, but then decided to use extend it for pi.dev. I now have a dedicated docker image for pi.dev, that only contains Yamlet plugin, and whenever I work on spec I spin it up.

The end result is a .yaml file that easily works in git + git diff, so that I can then proceed with the technical specs-

Yeah Im pretty sure gambling the HN frontpage is just a small transaction away. How else would this, sorry for my english, half finished, vibe coded bull get so far

I like to point out that if all people do is copy paste from an AI, then I can advocate to replacing them with AI directly. Usually wakes people up.

No? You work on something and finish it. At most I have 2-3 feature branches open. If none are in review, I have commits in them with current work. Maybe I use the stash 2-3 times a year when I am heavily experimenting with different implementations.

Interesting how posts like these seem to be catapulted to #1 spot so quickly

Along with a number of posts praising the "website design". Besides novel designs (I think of Acko.net) it's not often I see comments on that here.

As the original poster and creator I am actually a little surprised, too!

I've been working on and off on flagforge.dev

I have always hated how we deal with feature flags; often times as a dev I implement it, but then business decides how the flags are used, which comes back to us having to set them. I want to make this something that can be used by both easily, and priced fairly - looking at how much others charge for a feature flag SaaS is insane.

Currently I'm in heavy testing by myself, as I am quite worried about issues (that's just my personality), but yeah. It started as a normal side project for me to train a bit, then I added a bit of agentic coding to it to also learn it, and now it's here.

Great one, I've been hoping for a kagi alternative in the EU, will check it out :)

Thanks! Please let us know what you think is missing or needs to improve!

Not really. If pay decreases it's because you're not required anymore or less, which is contrary to what has been shown. IF educating and enabling juniors etc. is not handled correctly, then senior pay will explode, because whilst they are much more efficient, their inherent knowledge is required to produce sustainable results.

> If pay decreases it's because you're not required anymore or less

Not necessarily, there are many factors at play here which are downplayed. The first one is education: LLMs are going to significantly improve skill training. Arguably, it is already happening. So the gap between you and a middev will get narrower. At the same time, candidates who can be as good as you will increase.

While you can argue that you possess specialised skills that not many do, you are unlikely to prove that under pressure within a couple of hours and certainly not to the level where you can have late 10s level of negotiating power imo.

At the end of the day, the market can stay irrational longer than you can continue refuse to accept a lower offer imo. I believe there will be winners. But pure technical skill isn't the moat you think it is. Not anymore.

Green users are recently created accounts.

Skills are literally technical documentation for your project it seems. So now we can finally argue for time to write doc, just name it "AI enhancing skill definitions"