I've been using superpowers:brainstorming skill with claude for a while and have found the process to work well for me. I generally spend 10-30 minutes typing my requirements and ideas for a feature and then I feed it into brainstorming skill which asks questions to fill in the gaps before starting to build things. I don't bother reading the spec it creates. I have been consistently happy with the results.
However, since everything changes every 5-minutes now; I am curious what is now a better process than using superpowers? What works for you?
I am getting really good results from claude. We have a 22-year old legacy system. The system is stable, but had issues as all legacy systems do. Claude has been great for modernizing the codebase, updating dependencies, auditing security, and rapidly adding new features. It has worked well with existing code style and patterns. Sometimes it is a little off-track, but overall it is pretty amazing.
When implementing new features or making large refactoring changes; I use the superpowers:brainstorming skill. That has consistent process which has worked really well. I alway review the code before merging, but most of the time there are few issues to correct.
I don't do 95% coverage, but I have increased it from 65% to about +80% and that is sufficient.
Lately, I've seen a couple responses to my comments which inquire about metrics. They seem strange and I wonder if they are bots. I just noticed this inquiry is from an account that is 11 days old. How does one benefit by adding bot comments in a forum like this?
I can understand normies being AI skeptics, but at this point it is a pretty insane position for any developer/nerd. It's like being schizophrenic and completely out of touch with reality.
Asking for proof of improvement is a very rational reaction to the last three to five years of constant AI hype cycling.
It's a simple question that seems to make AI cheerleaders really mad.
"How are you measuring improvement."
I use AI for coding every day and see it fail all the time, I'm a seasoned developer and early tech adopter just like everyone else on HN. I'm still very skeptical because of how often these systems just miss. It's gambler's ruin on a very large scale, we remember the hits and forget the misses.
What kind of metrics are you looking for? No one answers that question. It's like asking for clinical trial results about whether or not a parachute works and if you should use one when jumping out of a plane.
I don't expect AI to be perfect. Are people perfect? I leave room for corrections, but at this point they are usually minor and acceptable.
If this is true, then you are not saving a lot of time. Because most of the time is spent evaluating various options and ways to implement the functionality. Even when you are reviewing, you ll have to do that. (With LLMs, this is even more feasible, because now you can actually implement some of the variants, and evaluate them).
But on the other side, you are saving from typing the code. So if you are really reviewing everything, then you are not saving much time. The alternative is that you settle for some local maximum during each review, that in long term won't necessarly translate to a globlal maximum or even a global "good enough" position...
I'm saving time. It's implementing features in a few hours that would have taken me weeks to implement. I can review code much faster than I can create well reasoned, implemented and tested solutions.
This faux socratic method flavor of AI skepticism is so cringe at this point, and reeks of your own insecurity in your beliefs. We get it, you don't think AI is useful for the type of coding people claim it is. Why on earth would anyone try and convince you at this point?
This is not a binary thing. It could be useful at the same time detrimental in some manner. Look at smartphones. I am just raising the possibilities if one use LLMs indiscriminately to generate code.
Probably because it's utterly pointless to try and convince someone in 2026 that LLMs are useful for coding. Why waste time trying to talk to someone with their head in the sand?
I can understand those pain points, but they don't really bother me. I'm on my 3rd oura ring now and I still like using it. I don't go crazy with the information, but I really do try to get a good night sleep and oura lets me know when I do not. I got married a few years ago and I usually don't like wearing my wedding band around because I worry about loosing it. So I wear the oura ring in place of the wedding band. I take the ring off when I lift weights and do some other things.
I also wear a garmin watch that I mostly use for tracking mountain bike rides, hikes, etc. So I'm just a geek who enjoys gadgets and the data they provide. I like that both have very long battery lives. I would love all the features of an apple watch, but I don't want to be charging it every day.
Does your body not do the same? I really want to like sleep tracking as a feature, but I find that my body already has built in sleep tracking, and the tech version doesn't give me any actionable advice on how to improve my sleep that my biological sleep tracker isn't already giving me.
I could see how it would be helpful for people who don't necessarily know they have sleep apnea. I also recently read about someone who quit smoking weed, and their sleep tracking changed dramatically while they detoxed. With my Garmin, that and the Stress tracker usually let me know 1-2 days ahead of time that I'm getting sick, before I even feel it. My sleep quality plummets, and my physiological stress is at like 80%+.
The only thing my body tells me in the morning is.. "dude coffee now". I have no problem falling asleep, but I usually wake up pretty early. I feel ok with 6 hours of sleep, but I don't think that is optimally healthy. I don't worry about it a lot, but I look at sleep duration, heart rate, hrv, and sleep debt. The details motivate me to pay attention to sleep hygiene. If I ignored it, I don't think I would sleep enough.
I would expect the impact of cpu limits to be different between K8s providers. I have only used memory limits. If I had a pod that tended to be very cpu intensive, I would schedule it on it's own node group.
However, since everything changes every 5-minutes now; I am curious what is now a better process than using superpowers? What works for you?