An agent needs reliable structured data from the model at each step. What do you use?
Why A is the answer
A schema for the output, and a check on it
Structured output plus validation stops bad data moving on.
In detail
- A schema shapes output, and a check confirms it.
- Reject bad data early.
Use case
An illustrative scenario, not a real incident.
Each step must produce valid data for the next, so output follows a schema and is checked.
See it in different ways
sequenceDiagram participant M as Model participant A as Application participant T as Tool M->>A: Tool name and arguments A->>A: Validate against the schema A->>T: Run, with an idempotency key T-->>A: Short result or clear error A-->>M: Result as context
flowchart LR S0["Schema"] --> S1["Model output"] --> S2["Check"] --> S3["Next step"]
Key point: Structured output plus validation.
Why the other options are wrong
- B. A request in the prompt, and trust in the reply
- A reply is not guaranteed.
- C. A longer temperature range across the steps
- Temperature does not shape data.
- D. A separate retrieval index for each output field
- An index is about search.