AI for QA / Scenario
Repair the Failing Test-Case Prompt
Investigate weak output from a test-generation workflow.
- Difficulty
- Hard
- Format
- Scenario
- Points
- 200
- Estimate
- 15 min
// MISSION BRIEF
Your Mission
A teammate's prompt for generating test cases keeps producing shallow, duplicate, happy-path-only results and blames the model. The prompt is the problem. Read it, find the specific defects, and choose the fixes that actually change the output.
// FIRST CONTACT
Battle teaser
First artifact
The current prompt
Which concrete defects in the PROMPT explain the weak output? (Select all that apply.)
- ANo context: the feature's actual rules (lockout, rate limit, enumeration, password policy) are never given to the model
- BThe model is simply too weak for this task and no prompt could help
- CNo output format: 'some good test cases' produces titles with no steps or expected results
- DNo coverage direction: it never asks for negative, boundary, or security cases, so it defaults to happy path
Answers, scoring, hints, and the full battle stay sealed.
// SKILL TAGS
prompt-engineeringtest-case-generationscenarioai-for-qa