Problem
Scenario
You are building an LLM agent that calls external APIs to complete user tasks. In testing, the agent repeatedly invents parameter values or passes fields that are not in the tool schema. The behavior is intermittent, so simple prompt tweaks have not fixed it.
Question
If an LLM agent repeatedly hallucinates parameters when calling an external API tool, how would you systematically debug and resolve this issue?
You are practicing as a guest. Sign up free to get your answer graded with AI feedback. Your draft stays right here.
