Đại học Kinh tế Quốc dân – National Economics UniversityMICRO4ASIA
🏆 Arena
Sign inSign up
MICRO4ASIAThe six micro-credentials

Courses are delivered on the Limio learning platform. Limio

← Prompt Engineering

You are viewing a public preview

Sign in to save your progress, do assignments, join discussions and get feedback.

Sign in
Module 2 — Design: Prompts and Prompting Strategies

Period 4 · Prompt Anatomy

Print / Save as PDF
Lecture video — Period 4: Prompt Anatomy

The five parts of a prompt and how OpenAI, Google and Anthropic organise the same requirements.

Flashcards · Prompt anatomy

10 words
WordMeaningExample
  • Prompt anatomy

    A practical breakdown of a prompt into functions. It is not a compulsory sentence order.

  • Instruction

    The action you ask the system to perform.

    Sum Net_Sales_Amount for each Store_Region.

  • Context

    The purpose and audience: who uses the answer and for what decision.

    A retail manager reviewing regional activity.

  • Input

    The source material or data. If the model cannot access it, the input is missing.

  • Output specification

    The required structure and content of the result.

    A four-row table ranked highest to lowest, plus two observations.

  • Constraint

    A defined boundary on the answer: what to include or avoid.

    Do not invent currency or infer profit.

  • Role (optional)

    Sets the assistant's purpose or tone. It cannot make a numerical answer accurate.

  • Provider layouts

    OpenAI, Google and Anthropic group the same needs differently; none is mandatory. Compare layouts by testing outputs against one rubric.

  • Does this sentence earn its place?

    Keep a sentence only if removing it would change an observable property of the answer.

  • Exact field names

    Naming real columns (Net_Sales_Amount, Store_Region) keeps the task checkable; asking for fields that do not exist invites invention.