How to actually tell AI what you want, so you're not fighting it all day.
Nina Brenes··6 min read
- MASTER stands for Markdown, Act as a, Specific, Thoughtful about Threads, Examples, Refine.
- The biggest lever is being Specific: give context and success criteria, like briefing a very smart intern.
- Refine means editing the original prompt until it becomes a reusable asset, not a one-off chat.
- MASTER is AMP's prompting framework (runonamp.com), used in Nina's AI operations practice.
The MASTER Method is a six-part way to tell AI exactly what you want, so you get a usable answer the first time instead of fighting it all day: Markdown, Act as a, Specific, Thoughtful about Threads, Examples, Refine. And yes, people keep telling me “prompting” is a trick that will disappear. Yes and no. The weird hacks will die. But knowing how to tell AI what you want, clearly, will keep mattering until AI can read our minds. It matters more than most people think: in a Udacity survey of 2,000 professionals, 3 out of 4 workers said they regularly abandon AI mid-task, most often because the output lacks accuracy or quality (Udacity, 2025). That is not the AI failing. That is a briefing problem, and it is fixable.
MASTER is AMP's prompting framework (formerly The AI Exchange), where I certified as an AI Operator, and it is still the one I use today. It is not about tricks, it is about fundamentals for controlling what AI generates. Six pieces.
Markdown, Act as a, Specific
M, Markdown: AI was trained on structured text, so it likes instructions with hierarchy, headers and lists. A, Act as a: if you start with “act as a world-class journalist,” AI primes itself to predict better answers. S, Specific: the most important one. AI does not read minds. Give it context (like briefing a very smart intern) and success criteria (what “done well” means: length, format, reading level).
Threads, Examples, Refine
T, Thoughtful about Threads: one task per conversation. If you mix topics in the same thread, AI gets confused by conflicting context. E, Examples: instead of only telling it what you want, show it examples of what good looks like, and it will imitate them. R, Refine: do not keep piling on follow-up messages. Go back to the original prompt, edit it, and run it again. That is how a prompt becomes a reusable asset instead of a one-off task.
- Markdown: structure it with headers and lists.
- Act as a: give AI a role.
- Specific: context plus success criteria.
- Threads: one task per conversation.
- Examples: show, do not just tell.
- Refine: edit the prompt until it is an asset.
MASTER does not live alone
It is the piece that goes inside each step of a playbook. Both frameworks, MASTER and playbooking, are from AMP (runonamp.com); together they are how you move from chatting with AI to delegating processes that actually run.
FAQ
What is the MASTER method for prompts?
MASTER is a prompting framework from AMP: Markdown, Act as a, Specific, Thoughtful about Threads, Examples, Refine. It is a set of fundamentals for communicating with AI so it reliably produces what you want.
How do I write a good prompt for ChatGPT or Claude?
Structure it with markdown, give the AI a role, be specific with context and success criteria, keep one task per thread, add examples, and refine the original prompt instead of chatting endlessly. That is the MASTER Method.
What is the most important thing when writing a prompt?
Being specific. AI cannot read your mind, so give it context and a clear definition of success, the way you would brief a smart intern who knows nothing about your situation. That single move improves most outputs.
Why should I rewrite the prompt instead of continuing to chat?
Because editing the original prompt turns it into a reusable asset you can run again and hand to your team, while endless follow-ups stay a one-off. Refining is what sets you up to automate the task later.
Nina Brenes
AI partner for purpose-led, human-first founders
Certified AI Operator (The AI Exchange) and Anthropic Claude Partner. Nina spends her own time in AI conferences and daily practice so the founders she works with don't have to.
Certified AI Operator · Anthropic Claude Partner
Connect on LinkedIn