Using the GitHub Copilot SDK for Java

GitHub has introduced the Copilot SDK for Java, a framework-agnostic client library that allows enterprise Java developers to programmatically interact with GitHub Copilot from their applications. The SDK supports annotations, virtual threads, and Jakarta EE environments, eliminating dependencies on specific frameworks like Spring AI or Langchain4j. It enables developers to create Copilot agent sessions, register tools, send prompts, and receive structured responses directly from Java code, simplifying AI integration in server-side applications.

The SDK includes features such as the @CopilotTool annotation for defining AI tools, which the model can invoke during processing. Tools can be registered via annotations or lambda wrappers, with support for overriding built-in tools when custom behavior is required. The SDK also provides fine-grained control over system messages, allowing developers to customize prompts while preserving safety guardrails. A sample Jakarta EE 11 application demonstrates these capabilities, showcasing a real-estate lead-management agent pipeline that processes customer inquiries concurrently using virtual threads.

The sample application highlights the SDK's event-driven architecture, where tool calls, results, and assistant messages trigger real-time updates via Jakarta WebSocket. This enables developers to monitor agent progress through phases directly in the browser. The architecture leverages virtual threads for efficient concurrent processing, with each inquiry handled by an isolated Copilot agent. The SDK's experimental annotation-based tool API requires specific Maven configuration, ensuring compatibility with enterprise Java standards.

The SDK supports CopilotClientMode.EMPTY for direct CLI interaction, and its event system provides granular control over agent behavior. Developers can use pattern matching to handle specific event types, while a custom executor ensures tool callbacks maintain container context. The release emphasizes open standards and framework neutrality, positioning the SDK as a flexible solution for integrating AI into enterprise Java applications without vendor lock-in.