The Bill Guard

# The Bill Guard

Your AI runs its most expensive brain on your dumbest jobs — and you only find out on the invoice.

Renaming a file. Sorting a list. Formatting some json. None of it needs a genius — but the model doesn't know that. It just grabs the biggest, priciest brain for every task, every time, and quietly bills the whole trivial 90% at premium rates. The leak isn't that your AI is expensive. It's that it never asks which brain the job actually needed.

The Bill Guard is the one rule that plugs it: send the cheap 90% to a small model, and reserve the expensive brain for real judgment. Same jobs, same output — one line decides who runs each one. On an example month of mixed work, that drops the bill about 80%, because you stop paying premium for the reading, sorting, and checking that never needed it.

Almost everyone's setup has the same silent leak: the expensive model is the *default*, and the cheap one is a thing you'd have to go out of your way to choose. So everything — including work a calculator could do — lands on the top tier by accident. The Bill Guard flips it: cheap by default, expensive on purpose. The top model becomes something you reach for deliberately for a judgment call or a plan, never the thing you fall back to without noticing.

• `ROUTING-RULES.md` — which of your three tiers runs which kind of job (read/classify → cheap, build/change → mid, judgment/verdict → top), plus the two guard-rules that stop bill creep.
• `cheap-default.config.md` — a drop-in block for your system prompt / CLAUDE.md / agent config that makes the cheap model the default and forces an explicit override to reach the expensive one.
• `routed-job-example.md` — one real job routed end to end, so you can watch the rule decide.
• `SETUP.md` — the three steps to wire it in.

It's tier-agnostic: map CHEAP / MID / TOP onto whatever you have — Claude Opus/Sonnet/Haiku, GPT-5/mini, Gemini Pro/Flash. The names don't matter; the tiers do.

1. Map your three models cheapest → priciest and label them CHEAP / MID / TOP.
2. Paste the cheap-default block into your system prompt / CLAUDE.md, tier names filled in. This alone plugs the biggest leak.
3. Route sub-work down: helpers are mid-or-lower, read-only checkers are cheap, and always pass the model choice explicitly — never inherit.

Then grep your own setup for the anti-pattern: *the top model doing mechanical work.* If your priciest model is reading, sorting, or classifying — route that step down and watch next month's bill.

## Get it now

The BILL keyword retires at 4pm today. Comment BILL on the reel and the kit lands in your DMs — the routing rules, the cheap-default config, and one job routed start to finish. Pay the setup once; the trivial 90% never touches the expensive brain again, and the bill stays capped every month.

_Built by SharpenDaily — your AI, sharpened._