The Role of LLM-Powered Agents in Automating Test Processes CT-GENAI Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   What is the primary role of a human tester when working with an LLM-powered agent?

   • ATo manually perform all test executions while the agent only generates ideas
   • BTo supervise the agent's outputs and validate that they are correct and relevant
   • CTo write all test cases from scratch without using the agent's suggestions
   • DTo monitor the agent's hardware usage and ensure it does not overheat
   Reveal Solution Discussion Select an answer first
2. 7expert · hard

   A team wants an LLM-powered agent to generate unit tests for a new microservice. The agent produces a large number of tests, but a code-coverage analysis shows that many critical branches are not tested. The team has limited time to review the generated tests. What is the BEST way to improve the agent's output without a full manual rewrite?

   • ARun the generated tests, feed the coverage report back to the agent, and instruct it to generate additional tests for the uncovered branches
   • BAsk the agent to generate more tests in general, with a prompt to 'be more thorough'
   • CIncrease the temperature setting of the LLM to encourage more diverse test generation
   • DManually review the generated tests and write the missing branch tests by hand
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A team is designing an LLM-powered agent to automate API testing. The agent needs to read the OpenAPI specification, generate test payloads, execute the API calls, and parse the responses to verify the status codes. Which set of components is REQUIRED for this agent to function?

   • AA separate UI automation tool and a browser emulator to test the API through the frontend
   • BA fine-tuned model that has only been trained on the company's internal API documentation
   • CA database to store all historical test results and a reporting dashboard for the QA team
   • DAn LLM for generation and analysis, a tool to make HTTP requests, and a parser to interpret the API responses
   Reveal Solution Discussion Select an answer first
5. 9application · medium

   A team wants an LLM-powered agent to generate test cases from user stories. The agent is given a user story and is expected to produce both positive and negative test scenarios. The team notices that the agent only generates positive test cases. What is the MOST likely reason for this, and what is the BEST fix?

   • AThe user story is too vague; the agent needs more requirements to generate negative tests
   • BThe prompt does not explicitly instruct the agent to consider negative scenarios; update the prompt to require both positive and negative cases
   • CThe agent's tool for executing tests cannot handle negative test cases; replace the tool
   • DThe LLM is not capable of understanding negative test scenarios; use a different model
   Reveal Solution Discussion Select an answer first
6. 10expert · hard

   A QA manager is evaluating an LLM-powered agent to generate test cases for a safety-critical medical device application. The team has strict regulatory requirements for traceability: every test must be linked to a specific requirement ID. The agent generates many useful tests, but it sometimes invents requirement IDs that do not exist. The manager wants to adopt the agent while maintaining compliance. What is the BEST strategy?

   • AFine-tune the agent on the full requirements database so it never generates a non-existent ID
   • BUse the agent to draft test cases, but require a human tester to validate and correct every requirement ID before the test is approved
   • CConfigure the agent to only generate tests for requirements that it can find in the provided documentation
   • DDisable the agent's test generation feature and use it only for test execution and result analysis
   Reveal Solution Discussion Select an answer first