Test-case generation from stories
The problem
Writing positive, negative, boundary and edge cases for every story is slow and repetitive.
Workflow
- Story IDQA-482
- FetchJira tool
- GenerateLLM + team template
- Structuredoutput JSON → Excel / Xray
- QA reviewedit, approve
Real-world example
Story: 'Apply coupon at checkout.' Output: 14 cases — valid coupon, expired, case-sensitive code, coupon on sale items, minimum cart value boundary (₹499 / ₹500), two coupons, coupon after removing items. QA deletes 2 duplicates and adds 1 business rule.
Starter prompt
Generate test cases for the story below. Cover positive, negative, boundary and edge cases. Return
JSON array: [{id, title, preconditions, steps[], expected, priority, type}]. Use only rules stated
in the story; list assumptions separately.
STORY: {story}Tools
Jira, LLM (any), Excel / TestRail / Xray
Human check
Review for invented rules and duplicates; never import unreviewed cases.
Measure success
Time per story; % generated cases kept after review; defects found by AI-added cases.