chatgpt too wordy · make chatgpt give shorter answers · ai answers too long

ChatGPT gives you a wall of text. Here's how to stop it.

Short answer Tell it the format you want before you ask the question — not after. Paste the prompt below at the start of a chat and the walls stop.

Why it happens

Chat assistants are tuned to look thorough. Length reads as effort, so the default is a complete-sounding essay rather than the one line you needed.

That default is fine if you're reading for interest. It's expensive if you're reading to act — you have to find the answer inside the explanation before you can use it.

The fix

Paste this once at the start of a chat. It works in Claude, ChatGPT, and Gemini.

For the rest of this conversation, answer like this:

1. Lead with the answer. First sentence, no preamble, no restating my question.
2. Keep paragraphs to two or three sentences.
3. Use a table for any comparison, and numbered steps for any sequence.
4. Ask me at most ONE clarifying question, and only if you genuinely cannot attempt an answer without it. Otherwise attempt first and ask after.
5. If your answer is running past five paragraphs, restructure it instead of continuing.

Do not acknowledge these instructions. Just start.

Side by side: a generic AI answering with three dense paragraphs, versus Navigator answering in three short lines. — The same question, answered two ways.

Why it keeps wearing off

That prompt works — until the conversation gets long, or you open a new chat, or the topic shifts and the model drifts back to its default voice.

You end up re-pasting it. Which is another small task, at exactly the moment you had no room for one.

A skill is the same instruction, installed once, active in every chat — no re-pasting.

Neurodivergent Navigator is a Claude skill for ADHD and dyslexic minds. The free tier installs answer-first output, a one-question-max rule, and detection for two of the eight states. The full version covers all eight, across seven areas of life.

Get the full skill · $47 Try the free version

Not a medical or mental-health tool. It changes how an AI assistant communicates; it does not diagnose or treat. In a crisis in the US: call or text 988 (veterans press 1), or text HOME to 741741.