Spring AI AgentCore

Deployment

The SDK supports two deployment models:

• AgentCore Runtime — Fully managed: package as ARM64 container, push to ECR, create a runtime. Scales to zero, pay-per-use.
• Standalone — Use any module independently on EKS, ECS, EC2, or on-premises.

See examples/terraform/ for infrastructure-as-code with IAM and OAuth2 authentication.

Building the examples

The examples/ tree is a separate multi-module Maven build that depends on the AgentCore modules. Run mvn -DskipTests install at the repo root first so the examples can resolve the 1.1.0-SNAPSHOT artifacts locally, then:

mvn clean verify -f examples/pom.xml

Each module also has an AGENTS.md file providing context for AI coding assistants (project structure, conventions, key classes).

Quick Start

Add the BOM and the modules you need:

<dependencyManagement>
    <dependencies>
        <dependency>
            <groupId>org.springaicommunity</groupId>
            <artifactId>spring-ai-agentcore-bom</artifactId>
            <version>1.0.0</version>
            <type>pom</type>
            <scope>import</scope>
        </dependency>
    </dependencies>
</dependencyManagement>

<dependencies>
    <dependency>
        <groupId>org.springaicommunity</groupId>
        <artifactId>spring-ai-agentcore-runtime-starter</artifactId>
    </dependency>
    
    <dependency>
        <groupId>org.springaicommunity</groupId>
        <artifactId>spring-ai-agentcore-memory</artifactId>
    </dependency>
    <dependency>
        <groupId>org.springaicommunity</groupId>
        <artifactId>spring-ai-agentcore-browser</artifactId>
    </dependency>
    <dependency>
        <groupId>org.springaicommunity</groupId>
        <artifactId>spring-ai-agentcore-code-interpreter</artifactId>
    </dependency>
</dependencies>

Create an agent with memory, browser, and code interpreter:

@Service
public class MyAgent {

    private final ChatClient chatClient;
    private final AgentCoreMemory agentCoreMemory;

    public MyAgent(
            ChatClient.Builder builder,
            AgentCoreMemory agentCoreMemory,
            @Qualifier("browserToolCallbackProvider") ToolCallbackProvider browserTools,
            @Qualifier("codeInterpreterToolCallbackProvider") ToolCallbackProvider codeInterpreterTools) {
        this.agentCoreMemory = agentCoreMemory;
        this.chatClient = builder
                .defaultToolCallbacks(browserTools, codeInterpreterTools)
                .build();
    }

    @AgentCoreInvocation
    public Flux<String> chat(PromptRequest request, AgentCoreContext context) {
        String sessionId = context.getHeader(AgentCoreHeaders.SESSION_ID);

        return chatClient.prompt()
                .user(request.prompt())
                .advisors(agentCoreMemory.advisors)
                .advisors(a -> a.param(ChatMemory.CONVERSATION_ID, "user:" + sessionId))
                .stream()
                .content();
    }
}

record PromptRequest(String prompt) {}

This gives you a production-ready agent with streaming, conversation memory, web browsing, and code execution — deployed to AgentCore Runtime or standalone.

Examples

Spring AI AgentCore SDK

📖 Documentation

An open-source library that brings Amazon Bedrock AgentCore capabilities into Spring AI through familiar patterns: annotations, auto-configuration, and composable advisors.

Modules