Prompt Engineering · TRACK 01/05
Prompting Basics
Roles, examples, and structure: the fundamentals of talking to models so they listen.
// THE TRACK
01 · START HEREIntro to Prompt EngineeringUnderstand what prompt engineering actually is, why it survived better models, and which core skills transfer to every LLM you will ever use.BEGINNERZero-Shot vs Few-ShotUnderstand exactly what zero-shot, one-shot, and few-shot mean, and develop an instinct for when showing examples beats just describing the task.BEGINNERRole PromptingSeparate the folklore from the evidence on persona prompts: where assigning a role genuinely changes output and where it is just ritual.INTERMEDIATEXML, Markdown & DelimitersLearn the structural toolkit — tags, headers, fences, delimiters — that keeps a model from confusing your instructions with your data.INTERMEDIATEWriting a Good PromptWalk away with a repeatable checklist that turns vague requests into prompts a model can actually follow, with before-and-after examples.BEGINNERControlling Output FormatLearn the prompt-only techniques that pin a model to one output shape, so you can parse its answers reliably without bolting on structured-output APIs.BEGINNERNegative & Constraint PromptsUnderstand why negative instructions like 'never apologize' often boomerang, and learn to rewrite them as positive constraints the model can actually obey.BEGINNERControlling Tone & StyleLearn to steer a model's voice precisely — formality, length, warmth, brand tone — using style anchors and example text instead of vague adjectives.BEGINNERGuidanceYou will understand how Guidance interleaves your prompt with loops, conditionals, regex, and grammars to control exactly what an LLM is allowed to generate at each step.INTERMEDIATE