Ask HN: What does your AI workflow look like?

[tadziokas]

With the recently released models and various open source tooling like harnesses or even solutions such as Graphify (is anyone actually using this?) or the death of OpenClaw I feel rather behind with regards to how I apply AI during my day to day engineering tasks.

The goal for this post is then therefore to act as a probe into what the public considers as standard/best practices when working with AI and maybe act as a sort of sharing space for articles / ideas you have found interesting or apply while working with AI.

I feel like this could be of great value (although maybe the approach I'm taking right now is not the most efficient) to see what the public thinks is working/is bogus since all of the recent hype trains and their subsequent crashes make me (and potentially you too) feel out of touch.

Clarifications: - by saying AI I refer to AI assistants such as claude, codex or similar.

[kingkongjaffa]

- I have a central folder I use with context, project folders, tasks.md files etc. - I have a skill like /plan-my-day that makes a nice html file with tasks list, status of my work and projects etc. - when I'm starting up a new idea I use matt pocock's `grill-me` skill, I dictate a lot with https://handy.computer/ text going straight into claude code window. - once I have a decent brain dump I figure out what to do next, whether it's code up a prototype or find some one to bounce ideas off and ask some questions (an internal person or a customer)

I'm always looking for some kind of novel way that people have found to use, and the highest ROI thing I've found is the concept of the grill-me skill.

Please share other workflows, ideas, skills etc.

Some advice:

Don't install a whole plugin of skills, they use way too many tokens at rest in your context, instead work in your own .claude folder often, and vendor copies of specific skills or write your own.