Prompt

What Is a Prompt?

A prompt is everything you send to an AI model before it generates a response. It can be as simple as a single question or as complex as a multi-thousand-word document containing a role description, detailed instructions, background context, worked examples, and the actual task.

The model sees nothing except what is in the prompt, which is why what you put in, and how you structure it, matters enormously.

Anatomy of a Prompt

Most API calls to an LLM have three parts:

• System message: sets the model's role, persona, and ground rules ("You are a helpful customer support agent for Acme Corp. Be concise and friendly.")
• User message: the actual request or question
• Examples (few-shot): optional worked examples that show the model the pattern you want it to follow

Together these form a structured input that shapes the model's entire response.

Why Prompts Matter So Much

LLMs are sensitive to phrasing. Asking "Summarize this" versus "Summarize this in three bullet points for a non-technical executive" can produce dramatically different outputs. A vague prompt invites vague answers; a specific prompt tends to produce specific, useful ones.

Prompt Engineering as a Skill

Prompt engineering is the practice of designing, testing, and iterating on prompts to get reliable, high-quality outputs from AI models. It includes techniques like:

• Chain-of-thought: asking the model to reason step by step before answering
• Few-shot prompting: providing examples of good inputs and outputs
• Role assignment: giving the model a persona to adopt

It is a learnable skill that pays off quickly, even for non-developers.