Is your prompt working for you? Use this checklist to improve it!
Did you give it context? Does it know your role, what you are doing and who the audience is? If you showed your prompt to a stranger, would they have enough information to help you?
Did you give AI a role? If you need specialized knowledge, specify it. Do you need a biologist? A historian? A librarian?
Did you specify format? How do you want to view the output? Do you want a list? A table? A three paragraph summary?
Did you set the tone? Is the default okay or do you need something special? Formal? Informal? Concerned? Caring?
Have you been specific enough? Have you included all the steps and details needed for the response you want? Can you give an example of what you want or don't want?
It is rare that you will stop after one prompt. Keep adding details and asking questions until you get what you want.
What are ten questions I should be asking about this topic?
Explain [topic] as though I am [beginner/intermediate/expert]. Start with a one-sentence definition. Then explain how it works, why it matters and give a real-world example. If there are common misconceptions, address them. Limit the explanation to [# words/sentences/paragraphs].
Summarize key findings and arguements about [topic]. Cover the main perspectives, areas of agreement and major points of debate. Organize these perspectives by theme. Highlight practical implications and flag claims that would need verification. Limit the summary to [# words/sentences/paragraphs].
What experts, institutions and organizations are experts on this topic?
List discipline-specific jargon and plain-language equivalents for [topic].
What is the original source of (claim)?
What are the author's credentials and the publication's reputibility, include a citation trail.
Score this source on currency, relevance, authority, accuracy, and purpose.
Cross-verify key statistics against original datasets or registries where possible.
List common knowledge versus citable facts for my topic brief.
Check the research methods of the study for transparency, sufficient sample size, statistical reporting clarity and any funding conflicts of interest.