What is a Prompt?
A prompt supplies a task, relevant information and constraints. Its wording affects the answer, alongside model capability, available evidence and tools. A clearer prompt cannot supply a missing fact.
Core Principles
Fundamental guidelines for effective prompts.
Be Specific
Vague prompts get vague answers. Include relevant details and constraints.
Show Examples
Demonstrate the format and style you want with concrete examples.
Provide Context
Background information helps the model understand your needs.
Specify Format
Tell the model exactly how you want the output structured.
Anatomy of a Prompt
The components that make up an effective prompt.
Optional role
Use a role when audience, responsibility or style needs clarification; it does not grant expertise.
Task Description
What you want the model to do.
Context/Background
Relevant information for the task.
Output Format
How you want the response structured.
Build and review a prompt
Keep the source fixed while adding task constraints and acceptance criteria.
Build a prompt from an inspectable source
This assembles text locally. It does not generate model answers or measure prompt quality. Try this prompt on your chosen model and check every claim against the source.
SourceThe community library opens Monday to Friday, 09:00–18:00. Membership is free for residents. Visitors may borrow up to four books for 21 days. The announcement says nothing about weekend opening.
Summarize the library announcement for a new resident.
Review criteria: hours unchanged; free membership limited to residents; four books / 21 days; no invented weekend hours.
Key Takeaways
- 1State the goal, relevant evidence and acceptance criteria.
- 2Add representative examples only when they resolve an observed ambiguity.
- 3Compare revisions on the same representative task set, with model version and settings recorded.
- 4Verify facts and outputs independently; prompt length is not a quality metric.