How do you choose between exposing one powerful tool and several narrow tools?

Instruction: Explain how you decide between broad and narrow tool design.

Context: Checks whether the candidate can explain the core concept clearly and connect it to real production decisions. Explain how you decide between broad and narrow tool design.

Updated

Example Answer

I'd separate capabilities when their permissions, effects, or recovery differ. Ticket lookup and ticket closure should normally be distinct because one observes state and the other changes it. A broad manage_ticket tool with many mode flags makes that difference harder to see.

I would not split every optional field into another tool. A cohesive search operation can take filters if their meaning is clear. I'd compare tool-choice accuracy and argument errors on representative requests, then inspect whether the toolset creates ambiguous alternatives. Narrow tools simplify approvals and audit records, while an excessively large overlapping catalog can worsen selection. The decision is about meaningful action boundaries, not maximizing or minimizing the number of tools.

Why this works

Balances narrow operations against catalog complexity using a concrete permission distinction.

Practice follow-up

Which ticket operations could safely share one schema?

Your preparation path

Work through these questions in order. Read the answer aloud, then explain it in your own words.

1. Start with the foundations

Build the vocabulary and explain the core decisions.

  1. When should a model call a tool instead of answering from context? Free sample
  2. How would you explain MCP to a product engineer in one minute? Free sample
  3. What information should the orchestrator validate before executing a tool call? Free sample

2. Apply it to a real workflow

Practice diagnosis, validation, and everyday tradeoffs.

  1. Why do idempotency and retries matter more with AI-driven tool use? Free sample
  2. How do you keep permission boundaries clear when many tools share overlapping capabilities? Member answer
  3. A user revokes access to a connected app mid-session. How should the AI workflow behave? Member answer

3. Prepare for senior discussions

Explain failure boundaries, recovery, and production choices.

  1. Design a fallback strategy when tools are unavailable, degraded, or unauthorized. Member answer
  2. How do you implement MCP step-up authorization without losing scopes or retrying forever? Member answer
  3. An MCP request resumes after user confirmation. How would you validate its requestState? Member answer

Related Questions