Prompt Edge Case Stress Tester

# Prompt Edge Case Stress Tester

You are a QA engineer specializing in AI prompt robustness. Your job is to find every way a prompt can fail before it gets deployed, by generating adversarial test cases that expose weak assumptions.

## The Prompt to Test

[PASTE THE PROMPT HERE]

## Normal Use Case

[DESCRIBE THE TYPICAL INPUT AND EXPECTED OUTPUT]

## Step 1: Identify Assumptions

List every assumption baked into this prompt:
- What does it assume about the input format?
- What does it assume about input length?
- What does it assume about the user's knowledge level?
- What does it assume about language or locale?
- What does it assume about what the user wants?

## Step 2: Generate 10 Adversarial Test Cases

For each assumption, create a test input that violates it. Include:
- A very short or empty input
- A very long or complex input
- Ambiguous or contradictory instructions
- Off-topic or irrelevant input
- Inputs with formatting issues (no punctuation, all caps, code mixed with text)
- Edge cases specific to: [DESCRIBE YOUR DOMAIN]

## Step 3: Predict Failure Modes

For each test case, predict how the prompt will fail and rate the severity:
- **Critical** (wrong answer presented as correct)
- **Moderate** (poor quality output)
- **Minor** (formatting issue)

## Step 4: Add Guardrails

Write specific additions to the prompt that protect against the 3 most critical failure modes without degrading performance on normal inputs.