Define what “done” means
Short, narrated lessons. Understand an idea, see an example, then try it yourself.
Illustrated lesson with synthetic narration. Example boards explain the workflow; they are not recordings of Codex running.
Transcript & lesson chapters
0:00 · Understand the idea
Define what “done” means. Acceptance criteria describe observable outcomes. “The form rejects an empty email and shows a readable error” is more useful than “make the form better”.
0:13 · See a concrete example
See a concrete example. Write three checks for a contact form: valid submission, invalid input, and keyboard navigation. Specify the expected result for each.
0:26 · Try it yourself
Pause the video and complete the exercise.
0:35 · Avoid this mistake
Avoid this mistake. Testing only the happy path. Include an invalid input and a recovery route.
0:44 · Check your result
Check your result. A person who did not write the prompt can decide pass or fail for each check. Open the lesson below, try the exercise, and mark it complete only after checking the result.
Lesson notes
Acceptance criteria describe observable outcomes. “The form rejects an empty email and shows a readable error” is more useful than “make the form better”.
Try it yourself
Write three checks for a contact form: valid submission, invalid input, and keyboard navigation. Specify the expected result for each.
Verify your result
A person who did not write the prompt can decide pass or fail for each check.
Official references
OpenAI · Prompting ↗Product interfaces and access can change. Use the linked official guide for current installation and account requirements.