04 prompts

Module 04: Prompts

Deep dive into prompt requests and response handling.

What You’ll Learn

• PromptRequest structure — session ID and content list
• PromptResponse and StopReason values
• Content types for prompts

The Code

PromptRequest takes a session ID (from Module 03) and a list of content items. The response includes a StopReason that tells you why the agent stopped generating — check this to know if the response is complete, truncated, or refused:

// Send a prompt with text content
var response = client.prompt(new PromptRequest(
    session.sessionId(),
    List.of(new TextContent("Explain ACP in one sentence."))
));

System.out.println("Stop reason: " + response.stopReason());
// Output: Stop reason: END_TURN

Stop Reasons

The StopReason tells you why the agent finished:

StopReason	Description
END_TURN	Agent finished responding normally
MAX_TOKENS	Token limit reached
REFUSAL	Agent refused the request
CANCELLED	Prompt was cancelled by the client

Understanding stop reasons helps handle different agent behaviors. END_TURN is the normal case. MAX_TOKENS means the response was truncated. REFUSAL may require rephrasing the prompt.

Source Code

View on GitHub

Running the Example

export GEMINI_API_KEY=your-key-here
./mvnw exec:java -pl module-04-prompts

Next Module

Module 05: Streaming Updates — receive real-time updates during prompt processing.