Prompt engineering

The practice of writing instructions that reliably steer an LLM toward useful, predictable output.

Good prompts add role, context, constraints, examples, and a clear output format. They reduce the model's guesswork. Practice with real tasks beats reading prompt cheatsheets.

See it in the glossary →