DIGITAL KITCHEN
Back to glossaryGLOSSARY · Concepts

Prompt Engineering

The practice of crafting prompts to get better outputs from LLMs. Includes structuring instructions, providing examples (few-shot), and chaining prompts. The cheap optimisation before fine-tuning. Modern frontier models reduce the need for elaborate prompting but it remains relevant.

How it works

Effective prompts in 2026 are typically: clear role definition in the system prompt, structured input formatting (XML tags or JSON), explicit step-by-step instructions for complex tasks, examples for non-obvious patterns, and explicit output format specification.

Example

A content production agent uses a multi-section system prompt: ROLE (you are a B2B SaaS copywriter), CONTEXT (brand guidelines), TASK (write a 200-word LinkedIn post about X), CONSTRAINTS (no banned phrases list), OUTPUT FORMAT (JSON with body and hashtags).

Related terms

System PromptToken

Need to actually use Prompt Engineering?

We build production AI systems that put these concepts to work. 30 minutes, we map your use case.