Every task an agent takes on can run on a frontier model. Most shouldn’t. The ladder I run sends each task to the cheapest lane that fits, and a checklist keeps me from reaching for the expensive model out of habit.
Chapter 1 · Picking the model
Every “ALWAYS” and “NEVER” in your CLAUDE.md is a rule you’re hoping the model remembers. I turn those hopes into hooks my model can’t forget, and you can prompt your own Claude to do the same for you.
Chapter 2 · Guardrails
A model reviewing its own work has blind spots that match its own training. I run every change that would be expensive to get wrong past a panel of rival vendors instead, because the disagreement between labs is a signal one reviewer can never produce on itself.
Chapter 3 · Checking the work
$400 1 hour
Book a callA paid one-on-one call with answers to your questions about using AI for projects, building your AI stack, or growing your business with AI. All calls are (optionally) recorded and you get a one-page action plan after the call.
How I get a model to write in my voice instead of the flat, hedge-everything register everything else gets. I build both setups live: one that learns a voice from a corpus of my own writing, and one that reads a draft back and names the AI tells in it.
“After speaking with Kai I have clear, actionable steps on how to grow my business. There was no fluff, no BS, no running of the clock. Just Kai asking the right questions to understand my business and goals, then Kai providing clear sound answers on how to move forward.”
“I decided to speak with Kai because of his level of experience and the overall vibe around his communication: no hyping, no income claims, clearly stating the price. … After the call, I have more clarity, more confidence, and actionable steps I can take with my outreach.”
“I wasn’t sure the $400 call would generate a Return on Investment for my business, but it did. As a result of the call, I have a better idea on tools, processes, and positioning improvements I can make, all of which have made my business easier, more sustainable, and more scalable.”
The tool I lean on hardest, and the one thing I recommend to everyone: fan a change out to rival models before you trust it. Two pieces, because the code and the doctrine go stale at completely different speeds.
94.8 billion tokens
The last 29 days. At API list rates: $36,802.