• QA team wants to accelerate analysis, case generation, and defect triage using AI.
• Test scope changes rapidly and manual drafting is too slow.
• Team needs consistent prompt patterns for repeatable QA outputs.

• Use AI to boost QA productivity while keeping human quality control.
• Produce faster drafts for test points, risks, and bug analysis.
• Establish a safe AI usage pattern with validation checkpoints.

1. Define task objective and non-negotiable quality constraints.
2. Provide structured input context (requirement, data, boundaries).
3. Use focused prompts by task stage (analysis, cases, review, reporting).
4. Validate AI output against requirement and risk model.
5. Correct hallucination, ambiguity, and missing edge conditions.
6. Save validated prompt patterns for reuse.

Input:

• Requirement excerpt + known risks + output format template

Output:

• Draft test points with traceability and risk ranking
• Reviewer notes on unclear assumptions requiring human confirmation

Input:

• "Generate all tests for this feature"

Output (problem):

• Overconfident generic list, missing system-specific edge conditions
• No human validation checkpoint

• Do not treat AI draft as final QA deliverable.
• Do not input sensitive production data without policy approval.
• Do not skip human review for high-impact decisions.
• Do not use one prompt for all task types.
• Do not hide uncertainty in AI-generated conclusions.