Prompt engineering: zero to fluent
Six lessons. You start with what a single prompt is made of and end with few-shot examples and a self-critique pass. No experience needed.
-
Guide
What a prompt actually is
Start here even if you have used a chat model for years - the three-part structure explains why some prompts work and others quietly do not.
-
Guide
Giving a model a role and constraints
Naming a role, a format and an exclusion turns an unpredictable answer into a repeatable one.
-
Guide
How to write a prompt that actually works
A four-point checklist to run any prompt through before sending it.
-
Guide
Few-shot examples: showing instead of telling
Two or three worked examples often beat a page of written rules.
-
Guide
Force a strict output contract
For prompts whose output a script will read, not a person.
-
Guide
Add a self-critique pass before the final answer
A second pass that checks the first one against specific, known failure modes.