Prompt Patterns & Techniques
Move past one-off prompts and learn the named patterns pros reach for — worked examples, chain-of-thought reasoning, strict output shapes, and delimiter guardrails.
Chain-of-thought: think step by step
Ask for the working, not just the answer On multi-step problems — arithmetic, logic, planning, anything with several constraints — models do noticeably better when you ask them to reason step by step before giving the final answer. This pattern is called chain-of-thought. The one-line trigger Often it is as simple as adding “Think step […]
Chaining: when one prompt is too many jobs
The prompt that does four things does none of them well A request like “read these fifty reviews, find the themes, rank them by severity, and draft a response to each” looks efficient. It is the most reliable way to get mediocre output. Quality degrades on every sub-task, and when the result is wrong you […]
Making the model check its own work
A second pass catches what the first one missed Models are markedly better at spotting a flaw than at avoiding it while generating. That asymmetry is useful: ask for the answer, then ask for a critique of the answer, then ask for a revision. Each pass is a fresh, narrow job. The three-pass shape Pass […]