prompt-engineering.siWhat is prompt engineering? A practical starting point
Learn what prompt engineering is, how to write a useful brief, and how to improve a weak prompt with a worked project-update example.
Read the guide →The field guideUnderstand the method.
Understand the method.
Improve the next answer.
Eight focused guides. Original examples, concrete checks, and a clear next step in each one.
Prompt engineering examples: before, after, and what to check
Compare weak and improved prompts for summaries, decisions, editing, and extraction. Learn what changed and how to check each result.
Read the guide →Few-shot prompting: examples that teach the right boundary
Learn when to add examples to a prompt, how to choose boundary cases, and how to avoid teaching accidental patterns. Includes a classification exercise.
Read the guide →Structured output prompts: JSON, tables, and validation
Design JSON and table prompts with explicit fields, types, and missing-value rules. Understand where prompting ends and application validation begins.
Read the guide →Research prompts that keep claims connected to evidence
Build research prompts with a clear question, source boundaries, claim-evidence tables, and uncertainty rules. Includes a reusable synthesis workflow.
Read the guide →Writing prompts that preserve meaning and sound clear
Write better AI editing briefs with audience, purpose, voice, and protected facts. Includes a worked update and a repeatable revision checklist.
Read the guide →Decision-making prompts: compare options without outsourcing judgment
Use AI to structure a decision with alternatives, constraints, uncertainty, and reversal conditions. Includes a phased-rollout example and review checklist.
Read the guide →How to evaluate prompts: a small test set and clear criteria
Test prompt revisions with ordinary, edge, and missing-data cases. Use a reusable evaluation table to track correctness, evidence, format, and regressions.
Read the guide →