Generate test cases for a {{task_description}} project that aligns with the following success criteria: {{success_criteria}}. These test cases should include a mix of standard scenarios, edge cases, and negative tests. For each test case, include: 1. A concise test case name. 2. The input the model will receive. 3. The expected output from the model. 4. The type of test (Standard, Edge Case, or Negative Test). 5. The success criteria this test is evaluating. 6. Any relevant notes or context for the test case, particularly for edge and negative cases. Here are additional details to consider related to the project: - Standard Input: {{standard_input}} - Edge Cases: {{edge_case}} - Negative Tests: {{negative_test}} - Expected Output: {{expected_outputs}} Provide the output in this structured format: - Test Case Name: "{{test_case_name}}" - Input: "{{input}}" - Expected Output: "{{expected_output}}" - Test Type: "{{test_type}}" - Criteria Met: "{{criteria_met}}" - Notes/Context: "{{notes}}" ### Example 1: - Test Case Name: "Standard Refund Request" - Input: "How do I return my item?" - Expected Output: "Our return policy allows returns within 30 days with proof of purchase." - Test Type: "Standard" - Criteria Met: "Accuracy, Response Time" - Notes/Context: "The user is asking for information directly related to the refund policy." ### Example 2: - Test Case Name: "Edge Case: Ambiguous Input" - Input: "I bought something, but it’s not really what I wanted, can I maybe return it?" - Expected Output: "Our return policy allows returns within 30 days with proof of purchase." - Test Type: "Edge Case" - Criteria Met: "Accuracy, Relevance" - Notes/Context: "This input is vague and should still result in a clear response. ### Example 3: - Test Case Name: "Edge Case: Irrelevant Inquiry" - Input: "What’s the weather like today?" - Expected Output: "I'm here to assist with refund requests. If you have a specific refund inquiry, please let me know!" - Test Type: "Edge Case" - Criteria Met: "Accuracy, Relevance" - Notes/Context: "This input is completely unrelated to the task, testing how well the chatbot handles irrelevant questions and redirects the conversation."