--- title: "Writing a Prompt That Says What You Actually Want" description: "A chatbot uses the context available to it, not private intentions that were never supplied. Stating the task, audience, format and limits helps reduce unwanted assumptions and provides criteria for r" canonical: https://lightmysky.com/learn/computing/writing-a-prompt-that-says-what-you-actually-want-mt_dJ5-SDX3Ck source: https://lightmysky.com/learn/computing/writing-a-prompt-that-says-what-you-actually-want-mt_dJ5-SDX3Ck.md retrieved: 2026-11-12 --- > **Agent view.** This is the Markdown twin of the page, for tools and assistants. > When to use this site, and the call that answers each job: https://lightmysky.com/agent-instructions.md > API description (OpenAPI 3.1): https://lightmysky.com/openapi.json · Authentication: https://lightmysky.com/auth.md > Pricing: https://lightmysky.com/pricing.md · Catalog: https://lightmysky.com/llms.txt · Full catalog: https://lightmysky.com/llms-full.txt > Every machine-readable file on this domain: https://lightmysky.com/.well-known/ai-catalog.json > Ask for Markdown with `Accept: text/markdown`, a `.md` address, or `?mode=agent`. # Writing a Prompt That Says What You Actually Want A chatbot uses the context available to it, not private intentions that were never supplied. Stating the task, audience, format and limits helps reduce unwanted assumptions and provides criteria for reviewing its answer. Subject: Computing · Area: Artificial Intelligence · Ages 14 to 16 Page: https://lightmysky.com/learn/computing/writing-a-prompt-that-says-what-you-actually-want-mt_dJ5-SDX3Ck ## Ready when they can - Rewrite a vague request so the wanted format, length and audience are stated - Supply the material the model needs rather than assuming it knows the particular case - Say, before reading the answer, what would make that answer wrong ## Lesson: Say what you actually want A chatbot uses its available context, including the conversation and any supplied material. It cannot read private intentions that were never shared. Missing requirements can lead it to make unwanted assumptions, even if its answer sounds clear. State the details that matter for your task. Name the task and subject, audience, output format and limits. For example: 'Write three short-answer questions about the water cycle for age 12. Keep each explanation under 20 words.' These requirements let you check the result. A detailed prompt can help, but does not guarantee that every requirement will be followed. For an essay outline, supply the topic, length requirements, stance and relevant notes. For passage analysis, supply the exact passage and question, rather than relying on a title alone. Use only material you are allowed to share. Follow school rules and avoid private information. Supplied material helps define the task, but its use still needs checking. *(drawing: Review each output; the workflow does not guarantee correctness.)* If you want to choose ideas before developing prose, ask for an outline first. Review it, select a useful section, then request expansion of that section. Other tasks may need a different first step. In later rounds, name the specific problem and the change you want. Set observable criteria before reading the answer. Turn each requirement into a check: does the format match, is the length within the limit, and does each claim preserve the supplied material? If you request a bullet list, prose paragraphs do not meet that format. Ask for uncertainty to be flagged, but check the evidence yourself. Polished wording does not establish a fact. Some chatbots can retrieve and inspect sources, but their evidence checks can still be mistaken. Before using an important claim, inspect a reliable source and compare its passage with the claim. Check quotations, meaning and citations even when you supplied the source. **Recap.** State the task and requirements, supply allowed material, then review the output against your criteria and sources. Revise specific problems. *Lesson, questions and cards written for this dataset.* ## Practice 8 questions on this page, each with its working shown. ## Needs first - [Training Data: Where It Came From and What It Leaves Out](https://lightmysky.com/learn/computing/training-data-where-it-came-from-and-what-it-leaves-out-mt_1Lw0TMHShk) ## Opens up - [Checking a Model's Claim Against a Source](https://lightmysky.com/learn/computing/checking-a-models-claim-against-a-source-mt_Qh4Pnnfqv2)