AIF-C01 Practice Question: A company uses Amazon Bedrock Agents to automate…

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  By embedding the API directly in the agent's prompt instructions

  Why it's wrong here

  Prompt text cannot supply the structured operation schema Bedrock Agents require to invoke a function; the model may describe a call without executing it. Action groups with OpenAPI schemas exist precisely to expose callable APIs, making prompt embedding the wrong mechanism here.

• ✗

  By creating a custom model that has been fine-tuned to call the API

  Why it's wrong here

  Fine-tuning alters a model's weights for style or domain behaviour; it does not register an invocable endpoint with the agent runtime. Action groups are the supported way to expose external APIs, so a custom model cannot satisfy the requirement to call the enrichment service.

• ✓

  By defining an action group with the API specification and a Lambda function

  Why this is correct

  Action groups bind an OpenAPI schema describing the API's operations to a Lambda function that performs the actual invocation. Defining the group this way gives the agent a callable interface, satisfying the requirement to expose the external enrichment API without bespoke orchestration code.

• ✗

  By configuring a Bedrock Knowledge Base with API documentation

  Why it's wrong here

  Knowledge Bases perform retrieval over indexed documents for grounding responses; they do not define callable operations. Action groups, backed by OpenAPI schemas, are what let an agent invoke external APIs, so this misplaces the API definition in a retrieval component.

Go deeper

Related to this question