A team is building a tool-use agent and wants Claude to always invoke a specific tool (a search function) rather than answering from its training knowledge. Which API parameter achieves this most reliably?

The tool_choice parameter with type: 'tool' and a specific tool name forces Claude to call that exact tool in its response, regardless of whether it could answer from training knowledge. This is the authoritative API-level control for ensuring tool invocation. Option B (temperature: 0) affects output determinism, not tool use behavior — it does not force tool calls. Option C (system prompt instruction) is less reliable than the dedicated API parameter; instruction-following can be inconsistent under adversarial inputs or unusual conversation contexts. Option D (single tool in array) doesn't force its use — with tool_choice: 'auto' (the default), Claude can still choose to respond in text even when only one tool is defined.