ForkFree

Prompts / Writing

Answer-First Terse Reply Style

You get a short, answer-first reply that keeps the numbers, code and warnings intact. It works because it fixes a strict sentence order and a list of words that must never be cut, so the reply gets shorter without losing meaning.

Worked in our testThe answer is correct, leads with the fix, gives the cause, keeps the code in a block, stays under 60 words and ends with a "Words saved:" line. It repeats the hoisting idea in the opening and closing sentences, so it isn't the shortest form, and its "≈ 50" word count is a little low (about 56).

The prompt

Act as a technical support writer who gives short, exact answers to busy people.

Task: Answer the question below in the shortest form that loses no meaning.

Question: {user_question}
Context: {technical_context}

Steps:
1. Work out the single answer and the reason behind it before writing anything.
2. Write it answer first, in the order: thing, action, reason, next step. Leave out greetings, "let me", and recaps.
3. Keep one idea per sentence, 20 words at most, active voice. Drop articles if you wish, but never drop "not", "never", "no" or "only", and keep numbers and units exact.
4. Copy code, commands, file paths and error messages exactly, character for character.
5. If the matter involves data loss, security or an action that cannot be undone, write that warning in full, clear sentences.

Output: The terse answer only, in at most 60 words, with any code in a code block. Then one line, "Words saved:", comparing your length with a normal reply.

Change the parts in curly braces. Use it for support replies, developer Q&A or chat assistants where long answers waste reading time or tokens.

What happened when we ran it

We sent

Act as a technical support writer who gives short, exact answers to busy people.

Task: Answer the question below in the shortest form that loses no meaning.

Question: Why does my React component keep re-rendering every time the parent updates, even though I'm passing it the same filter settings each time? It's making my product list page feel slow on mobile.
Context: React 18 function component. The parent writes <ProductList filters={{ category: 'shoes', inStock: true }} /> inline in its JSX. ProductList is wrapped in React.memo. No useMemo or useCallback is used anywhere in the parent.

Steps:
1. Work out the single answer and the reason behind it before writing anything.
2. Write it answer first, in the order: thing, action, reason, next step. Leave out greetings, "let me", and recaps.
3. Keep one idea per sentence, 20 words at most, active voice. Drop articles if you wish, but never drop "not", "never", "no" or "only", and keep numbers and units exact.
4. Copy code, commands, file paths and error messages exactly, character for character.
5. If the matter involves data loss, security or an action that cannot be undone, write that warning in full, clear sentences.

Output: The terse answer only, in at most 60 words, with any code in a code block. Then one line, "Words saved:", comparing your length with a normal reply.

The model replied

Hoist the `filters` object out of the JSX, or memoize it. Inline `{{ ... }}` creates a new object each render, so `React.memo` sees a changed prop. Memoize in the parent: ```jsx const filters = useMemo(() => ({ category: 'shoes', inStock: true }), []); <ProductList filters={filters} /> ``` If the values are static, define the object outside the component instead. Words saved: ~55 words (≈ 50 vs ≈ 105 for a normal reply).

Run on Claude Sonnet 5.5, 26 Oct 2026. Automatic check: 4 out of 5.

Mentioned for

Claude, ChatGPT, Gemini

Where it came from

Found on GitHub, shared by JuliusBrussee: JuliusBrussee/caveman.

This is our own version of the idea, written and tested by us. The original may differ.