TAG Prompt Framework
TAG = Task, Action, Goal. A three-part prompt structure for direct, unambiguous requests. Specify what you are working on (Task), what the AI should do with it (Action), and what the result should achieve (Goal).
When to Use TAG
Use the TAG framework when:
- You need a quick, direct result without complex persona or constraint setup.
- The task is single-step — summarize, rewrite, generate, translate, explain.
- You are producing short-form content — social posts, headlines, code snippets, or email subject lines.
- The request is clear enough that extra context would add noise rather than value.
When Not to Use TAG
- Complex multi-step workflows — use the RISE framework instead.
- Tasks requiring a detailed expert persona — use the ROLE framework instead.
- Strategy or planning tasks — where full background context matters, use COAST.
TAG Framework Template
Example: Code Task
Before (weak prompt):
After (structured prompt):
Example prompt:
I have a Python list of dictionaries, each with 'date' and 'title' keys. Write a function that sorts the list by date in ascending order. Return clean, readable code with inline comments explaining each step.
Example: Marketing Copy
Before (weak prompt):
After (structured prompt):
Example prompt:
Write a LinkedIn post announcing an AI-powered prompt optimization feature for SaaS teams. Target product managers and developers. Keep it under 200 words and end with a call to action to try it for free.
How to Write a TAG Prompt — Step by Step
- Define the Task. Describe what you are working on — the subject, the material, or the problem context.
- State the Action. Specify exactly what the AI should do — write, rewrite, summarize, analyze, generate, translate.
- Define the Goal. Describe what the output should achieve — the purpose, target audience, or success criteria.
Frequently Asked Questions
What does TAG stand for in prompting?
TAG stands for Task, Action, Goal. It is a three-part prompt structure that specifies what you are working on, what the AI should do, and what the result should achieve.
Is TAG good for ChatGPT?
Yes, TAG works well for ChatGPT for direct, specific requests. The three-part structure removes ambiguity and gives the model enough context to produce a focused response without over-engineering the prompt.
When should I use TAG vs ROLE?
Use TAG for simple, direct requests where you just need a clear task completed. Use ROLE when you need the AI to adopt a specific expert persona, apply detailed constraints, or produce output in a tightly controlled format.
Try the Prompt Generator →Also see: All AI Prompt Frameworks · AI Prompt Optimizer