PROMPTHERE GUIDE
한국어로 읽기AI prompt templates: a practical guide to writing reusable requests
A reusable prompt template separates the information that changes from the rules that stay the same. It is more than a completed request copied into another conversation. Clearly named placeholders help keep an old company's name or an example figure out of a new answer.
Updated · Prompthere
Start with five practical parts
Define the goal, input material, constraints, output structure and review criteria. The goal names the deliverable; the review criteria explain how to check it. You do not need a long paragraph for every part. Include the details that would change your judgment of whether the result is useful.
Give placeholders meaningful names
Several fields all called ‘content’ make a template difficult to use. Prefer names such as audience, source text, length and prohibited phrases. Tell the model to ask before proceeding when a required input is absent. For optional fields, state what should happen if you leave them blank.
Test complete and incomplete inputs
Try the template with enough material to perform the task, then remove a date or figure. Check whether the answer leaves the gap visible instead of filling it with a guess. Change one instruction at a time when revising, so you can compare what actually affected the result.
An example to make your own
A general template that asks before guessing
Replace the bracketed fields, then paste into your AI tool.
Goal: [deliverable] Audience: [reader] Input: [source material] Constraints: [length, scope, prohibited phrases] Output: [headings and structure] Review criteria: [factual accuracy, format, other checks] Ask questions before proceeding if required information is missing. Leave unsupported facts unresolved. After drafting, list any review criteria the draft does not yet meet.
Before you run the prompt
- Separate reusable rules from changing inputs.
- Define what happens when a required field is missing.
- Use review criteria that can be checked against the answer.
Frequently asked questions
Do I have to fill every template field?
Use the fields relevant to the task. Distinguish required information from optional context and define how to handle any important omission.
Do I need to study prompt engineering first?
You can begin with a small task and a clear review standard. This guide focuses on understandable instructions and checking results, rather than requiring specialist terminology.