The Korvai Blog
Every session starts loaded.
Notes on prompt-craft, context engineering, and building a system so you never start an AI chat from scratch again.
The workflow you run every week should be a skill, not a habit
You have a dozen tasks you do the exact same way every week, and you rebuild each one from memory every time. Here's how to turn a repeated workflow into a skill you run on demand, instead of a habit you carry in your head.
Read →Context, prompts, skills: the three things that actually change your AI output
Most advice about getting better at AI is a pile of disconnected tips. There's actually a system underneath, and it has three layers. Once you see them, you stop collecting tricks and start building something that compounds.
Read →Your best prompts are trapped in one person's head
On every team there's one person whose AI output is noticeably better than everyone else's. Their advantage almost never spreads, not because they're hoarding it, but because the thing that makes them good was never written down. Here's how expertise actually scales past one brain.
Read →You can't ask what you don't know to ask
AI rewards the people who already know what to ask — and that's the same expertise novices are using it to borrow. There's a name for this trap, and the fix isn't a smarter model. It's borrowed articulation.
Read →AI doesn't refuse to answer. It quietly guesses who you are.
The same question gets a different answer for a 35-year-old than a 60-year-old — and the model never tells you which one it assumed you were. The real problem with AI advice isn't that it's wrong. It's that its guesses are invisible.
Read →The $variable pattern: writing prompts your whole team can reuse
Most prompts are written once and thrown away. A small habit - pulling the specifics out into named variables - turns a throwaway prompt into a template anyone on your team can run. Here's how to do it well.
Read →Turn any prompt you've used twice into a template
The prompt you nailed last week is already lost. Here's how to stop rewriting your best work from memory and build a prompt library that compounds.
Read →Context engineering for people who aren't engineers
The reason your AI output is generic isn't your prompting. It's that the model has no idea who you are. Here's how to fix that without writing a single line of code.
Read →Every AI session is day one
You re-explain who you are every time you open a new chat. The blank-box problem is why your AI outputs feel generic — and what to do about it.
Read →Skills as a git repo you never clone
How Korvai serves a dynamically-generated git repository over HTTP so Claude Code auto-updates your skills as slash commands — no clone, no copy-paste, no manual sync.
Read →