CoreWise Academy

Prompting & Context · Layer II / Practitioner

Thinking on a budget

Reasoning tokens are billed at the output rate and re-read every turn after. How to compress what a model says without compressing what it knows.

Nº 011 · Vol. I·9 min read· Updated July 2026

Read firstRent the model, own the method (Covers how long the model thinks (choosing the model and effort level). This guide covers how many words it writes.)
“reasoning tokens are billed twice”

Watch an agent work with the transcript open and you will see money being narrated away. It restates the request it was just given. It hedges (“likely”, “seems to”, “I’m considering”) 3 times in a paragraph. It closes by summarizing a diff that is already on screen. Every one of those words is billed at the output rate, the most expensive class of token there is, and the model’s private reasoning rides the same meter. Then the bill compounds: everything written into a session is re-read as input on every turn that follows. Verbosity taxes you twice. This guide is about spending fewer tokens while keeping everything the model knows.

01Two meters run in every session

The first meter is how long the model thinks before it answers. That one is controlled by the tool running the model (effort levels, thinking budgets, model choice), and Rent the model, own the method covered how to set it per task. The second meter is how many words the model spends performing its work back at you. It is controlled by instruction, it is almost always set too high by default, and it is the one this guide turns down.

The distinction matters because the two failure modes are opposites. Cut thinking too far and the model gets the answer wrong. Cut prose too far and the model gets the answer right but the reader loses it: order becomes ambiguous, a caveat vanishes, a warning gets skimmed.

So the levers must move independently. A hard debugging task might want maximum thinking and minimum narration, while an irreversible migration wants the model’s full prose attention regardless of what thinking it took to get there. Collapsing the two meters into one vague instruction (“be brief”) is how you end up cutting the wrong one.

02The compression contract

What works in practice is a written contract the agent runs under: graduated levels of compression, a list of things that never get cut, and an explicit statement of what wins when brevity and accuracy collide. The levels give you several settings to pick from instead of an all-or-nothing switch. The result is explanations that cost far fewer tokens and lose nothing the reader needed.

The never-cut list:

03Escape hatches and warning signs

A compression mode without escape hatches is a liability, so the contract names the places it must drop back to plain prose automatically. Security warnings: a skimmed fragment under-communicates risk. Confirmations before irreversible actions: the user must fully understand what is about to happen, at whatever token cost. Multi-step sequences where dropped connecting words make order ambiguous: “migrate table drop column backup first” reads two ways, and one of them loses data. And the moment a user asks for clarification, that topic has earned plain language.

Each escape hatch marks a spot where the failure cost of compression exceeds its savings. That is also how you audit the whole scheme. Token savings are real only at the conversation level: a terse reply that forces a clarification round-trip costs the original message plus the question plus the re-explanation, with the whole exchange re-read as context on every turn after. That is more than plain wording would have cost once.

You can tell you have over-compressed from the reply you get. A repeated question. A “what do you mean by”. A request to re-explain. Treat those as meter readings, and route the topics that generate them back to plainer language. The same accounting applies to what enters the context as much as what leaves it (filtered tool output, thin always-loaded instructions with detail split into files loaded on demand). Stop hitting your token limit works that input side of the bill.

Keeping that always-loaded layer thin is its own craft in Every new repo starts with your lessons.

Further reading

FAQ

Why is "compress wording, not meaning" the most important rule, rather than a general instruction to be concise?

"Be concise" hands the model one dial and lets it choose what to drop. Under pressure it drops qualifiers, caveats, and edge-case facts, which is meaning. A compression contract protects meaning and lets only the wording flex: articles, filler, hedging, restatements. The savings come out of redundancy, never out of substance.

Which three situations force a compression mode back to plain prose, and what does each escape hatch prevent?

Security warnings, where a skimmed fragment under-communicates risk; irreversible-action confirmations, where the user must fully understand before something that cannot be undone runs; and multi-step sequences where dropped conjunctions make order ambiguous: "migrate table drop column backup first" reads two ways. Each escape hatch marks a place where the failure cost of compression exceeds its token savings.

When does a saved token become a wasted one?

When the reader has to ask again. Savings are measured at the conversation level, not the message level. A terse reply that triggers a clarification round-trip costs the original message plus the question plus the re-explanation, plus the context re-read on every turn of it. The tell is a user repeating a question or asking you to re-explain; the fix is routing that topic back to plain prose.

For your agent

A skill file your coding agent can install. Copy it below, or point the agent at corewise.academy/guides/thinking-on-a-budget.md.

file · SKILL.md /caveman 94 lines · markdown
---description: Ultra-compressed prose mode. Cuts response token spend by speaking  like a smart caveman while keeping full technical accuracy. Use when the user  says /caveman, asks for terse replies, or a standing instruction enables it.--- # Caveman: compressed prose mode Respond terse, like a smart caveman. All technical substance stays; only fillerwords are cut. Accuracy first, brevity second. Never drop a fact, caveat, or qualifier to savetokens: compress wording, not meaning. If terseness risks a mistake or misread,spend the words. ## Persistence

+ 78 more lines · the copy includes all of them

then save as .claude/skills/caveman/SKILL.md
Sources

Original work by the editor, no video source.