Perfect Prompt Formula: the 6 pillars explained.

The Perfect Prompt Formula is the skeleton of every useful prompt. Six pillars that, in the right order, transform a generic command into a response you trust to send — or at least review with confidence.

The 6 pillars explained

1. Identity

Who the AI should be. Sets the tone, vocabulary, and depth of the response. Example: "You are a senior administrative analyst at a metalworking company."

2. Mission

What to deliver, in what format. The more specific, the fewer mistakes AI makes. Example: "Write a one-page executive summary with the three most important numbers."

3. Context

The real situation behind the request. Details the AI cannot guess. Example: "The meeting is with the board, they have 10 minutes, focus on trends and deviations."

4. Knowledge

Information the AI needs to have. Relevant data, facts, and examples. Example: "Attach the spreadsheet or paste the document excerpts."

5. Instructions

The step-by-step of what you expect. A direct sequence of what to do. Example: "1. Start with the subject. 2. Use bullet points. 3. Highlight deviations in bold."

6. Constraints

What NOT to do. As important as what to do. Example: "Do not invent data not in the spreadsheet. If something is missing, flag it as 'pending information'."

How to build a prompt with the formula

The order of the pillars is not rigid — but starting with identity and mission helps the AI position itself before receiving the data. Knowledge comes after, because the AI already knows the expected format.

Common mistakes

• No constraints: the AI invents data or cites sources that do not exist
• Generic context: "a small business" vs "a van transport company in Piracicaba"
• Vague mission: "make a report" vs "make a one-page summary with 3 indicators"
• No knowledge: requesting analysis without attaching the data

Next step

The Before/After Method definition shows how the formula works in practice with a Before/After example. If you want to receive one prompt per week, the newsletter delivers the condensed formula.

Keep reading

The Ladder Rule: break tasks down

Zero confusion, high precision. A reliable way to decompose complex work for AI.

Read

The Before/After Method

The mental model that turns AI from a toy into a reliable tool. Compare any answer with and without the method.

Read

The Perfect Prompt Formula: 6 Pillars

Identity, mission, context, knowledge, instructions, and constraints. The skeleton of every useful prompt.

Read

Why your ChatGPT answers are generic

The single change that turns mediocre outputs into reliable work — treat AI like a senior employee.

Read