APE Prompt Framework: Action, Purpose, Expectation

Study guide

APE Prompt Framework: Action, Purpose, Expectation

APE is the beginner framework that makes you say why. Action is the task, Purpose is the reason you need it, and Expectation is the shape of the reply you want. Most people write only the action, so the AI optimises for the outcome it guesses rather than the one you meant. A single so that line fixes that, and it takes under a minute to write.

This course is for busy, non-technical professionals who paste prompts into ChatGPT, Claude or Gemini and want a usable answer on the first try. You will learn to open with one clear verb, attach the material it works on, name the reader and the decision in a short purpose line, and set the standard for the result. You will also see the marketing APE of Audience, Purpose and Execution, its cousin ACE, and when to switch to CO-STAR or RISEN.

What is a prompt framework?

A prompt is the request you type into an AI tool such as ChatGPT, Claude or Gemini. A prompt framework is a reusable structure for that request: a short checklist of the parts a good prompt contains, usually named with an acronym so it is easy to remember. Why it helps: the AI fills in whatever you leave out with generic guesses. A framework makes sure you include the parts that change the answer, such as who it is for, what you want to achieve and the format you need.

Frameworks make prompts repeatable: you fill the same parts each time, save the prompt as a template and share it with colleagues. They are not magic words: the value comes from the detail you put in each part. They help most when a request is complex or open to interpretation; a quick, simple question can stay short. There are well over sixty named frameworks, and most of them remix the same ingredients: role, task, context, audience, format, constraints and examples.

In this course you learn APE, which is a fill-in-the-slots framework: an acronym where each letter stands for one part of your prompt. It is best for: short operational asks where one outcome matters: do this, so that, give me that.

What APE is and why the Purpose slot matters

APE stands for Action, Purpose, Expectation. Action is the task, Purpose is the reason you need it, Expectation is the shape of the reply you want. The model optimises for the outcome you name, not the one you meant. A clear Purpose line steers the answer towards your real goal, while a vague one leaves the AI guessing.

Expectation controls format, length and level of detail. Without it you often get a correct but unusable answer, such as a long essay when you needed a table. APE is one of the easiest frameworks for beginners, alongside TAG. There is little to memorise, so you can use it on the first try.

A useful test after reading any answer is to ask whether it serves the purpose you wrote. If it does not, rewrite the purpose line before rewriting the whole prompt. Weak purpose lines use words like good or nice. Strong ones name a decision, a deadline, or the person who will read the output.

Action: one verb and the material it works on

The Action is the first letter of APE and the verb that tells the AI what to do. It should be the very first thing in your prompt, before any context or constraints. Stick to one of four everyday verbs: draft, list, summarise, or rewrite. These map cleanly onto most office tasks and leave little room for the AI to guess.

Always attach the material the action works on, such as a transcript, an email, or a set of notes. Without it, the AI fills the gap with invented content. A vague verb like handle, look at, or help with gives the AI no clear deliverable. Replace it with a verb that names the output you want to hold in your hand.

APE is designed for single outcomes. If your prompt contains two verbs joined by and, split it into two prompts and run them one after the other. This keeps each answer focused, and it makes the Purpose line for each prompt much easier to write.

Purpose: the so that clause

The purpose is the why behind your action. It starts with so that and names the real reason you want the output, such as a manager deciding in two minutes whether to join a call. Purpose controls length. When the reader has two minutes, the answer has to fit two minutes, so the AI trims the material to match the time you named.

Purpose controls focus. A manager deciding on a call cares about risk, cost and timing, so the AI leads with those and pushes other details down. Purpose also controls what gets left out. Naming the reason tells the AI which background, history and side details are safe to drop.

A clear purpose lets the model push back. If the action cannot deliver the purpose, a good model will say so instead of quietly producing something that misses. Keep the purpose short. One sentence that names the reader and the decision is enough. You are giving a direction, not writing a brief.

Expectation: the shape and standard of the result

Expectation is the third part of APE and describes the shape and standard of the finished result, such as five bullets, a decision at the top, and under one hundred words. It is what turns a vague request into a usable answer. Every Expectation has two halves. The include half names what must be in the answer, and the exclude half names what must not appear. Writing only the include half leaves the AI free to add things you did not want.

Use the rejection test. If you would send an answer back for something, that something belongs in Expectation. Common examples are a wall of text, invented numbers, a missing recommendation, or the wrong tone. Naming a length limit and a count is one of the fastest ways to control output. Five bullets and under one hundred words are easy to check, and they stop the model drifting into long paragraphs.

Expectation sits on top of Action and Purpose in the APE stack. Action is the job, Purpose is the reason, and Expectation is the standard. Removing it does not stop the AI answering, it just makes it guess the shape. Expectation also protects accuracy. Telling the AI not to include figures you have not given it removes a whole class of invented detail, which is far easier than spotting and correcting it later.

The other APE and ACE: putting the reader first

The marketing APE stands for Audience, Purpose, Execution. Audience is the exact reader, Purpose is the job the message must do, and Execution is how it is delivered, meaning format, length and channel. ACE stands for Audience, Context, Execution. It replaces Purpose with Context, the background and voice rules a writer needs, which makes it the stronger choice for content that must stay brand consistent across many pieces.

All three frameworks share one habit: the reader or the reason comes before the wording. Naming the person and the job first stops the model from writing generic text you then have to rewrite. Mixing the two APEs in a single prompt is a common error. If you write Audience and Execution but leave out Context, the model has no background to work from and the voice drifts.

Naming a crowd instead of a person is the quiet failure. A prompt that says everyone gets an answer for no one, while a prompt that names operations managers gets language that fits them. Execution is easy to forget because it feels obvious. Without it you get the wrong length, the wrong format or no clear next step, and the draft needs more editing than writing.

APE in your job: build the habit and know when to switch

The fastest way to learn APE is to rewrite prompts you already sent. Pick three from the past week, add a so that line, and compare the old answer with the new one. The difference shows you what the Purpose line actually does. APE has no role or context slot. That is a feature, not a flaw, because most simple tasks do not need one. When the answer depends on your role, your team, your rules or your audience, add a single line of context above the Action.

A prompt that keeps failing after several rewrites is telling you something. The task probably needs CO-STAR or RISEN, which add the role, audience, tone, steps or narrowing that APE deliberately leaves out. CO-STAR suits writing that must sound right for a specific reader, because it carries context, style, tone, audience and response format. RISEN suits multi stage tasks, because it carries role, steps and narrowing.

Notice the worked example. The nurse manager prompt worked without stating the setting, because the night shift notes already carried it. When your source material holds the context, you can leave it out. When it does not, add the line yourself. The three APE lines take under a minute to write. That speed is worth protecting. Reach for a bigger framework only when the task genuinely needs one, not out of habit.

Frequently asked questions