Podstawowe struktury promptów

A prompt is a set of instructions for the AI - like a recipe. The clearer the recipe, the better the result. Let's break a prompt into its core parts.

Good structure pays off: it prevents misunderstandings, sharpens accuracy, and keeps outputs consistent. The AI knows exactly what you want.

The instruction is the core command - what task to perform. Be direct and specific; vague language leaves the model guessing.

A plain instruction like "Summarize the following text" makes the model's job crystal clear. That directness is the whole point.

Context gives the model background or a scenario. It helps the AI grasp the situation so its answer fits your specific needs.

Context often sets a role or audience, like "You are a helpful assistant" and "for a 5th grader." Those framing words steer tone and depth.

Input data is the actual content the model works on to fulfill your instruction - text, a list, numbers, or any specifics it should process.

The input data is the raw material. In a summarize prompt, the text you paste in (often inside quotes) is exactly that data.

The output format says how you want the answer shaped - a bullet list, JSON, or "max 50 words." It keeps results consistent and easy to use.

Spelling out the output format, like "in exactly three bullet points," is vital when feeding LLM responses into an app that has to parse them.

You've learned about instructions, context, input data, and output format. Read the prompt below carefully.

"You are a helpful assistant. Summarize the following article for a busy executive. Provide the summary as a bulleted list of key takeaways.

Article: [Content of a fictional article about a new tech trend]"

Which part of this prompt specifies how the answer should be presented?

Recap: a strong prompt has four parts - instruction, context, input data, and output format. Master these basics for clear, predictable results.