Completion

SDK Usage

Learn how to evaluate this metric programmatically in the Trustwise SDK Documentation.

The Completion metric evaluates whether your AI agent fully and correctly completes the task requested by the user. It checks the thoroughness and accuracy of the agent’s response to the original query.

FAQs

What does a high Completion score mean?

It means the agent understood the user’s query and provided a full, detailed, and correct response.

When is Completion especially useful?

When your agent is expected to carry out multi-step tasks or generate long-form, detailed answers (e.g., recipes, instructions).

What if my agent partially answers?

Partial answers result in lower Completion scores, highlighting areas for improvement in understanding or generation.