The model keeps calling a write tool when a read tool would answer the question. How would you fix it?

Instruction: Explain how you would debug unsafe tool preference in an assistant.

Context: Tests how the candidate diagnoses the problem, chooses the safest next step, and reasons through recovery. Explain how you would debug unsafe tool preference in an assistant.

Official answer available

Preview the opening of the answer, then unlock the full walkthrough.

I would first make sure the read path is clearly the default for informational requests. If the model sees two tools that both sound plausible, it may choose the one with broader power simply because its description looks more generally useful.

Then I would inspect the interface...

Upgrade to view official answer

Related Questions