Google has officially announced the stable release of Android Studio Rabbit 2, marking a significant milestone for mobile development environments by bringing production-ready artificial intelligence tools and extended reality testing directly to developers’ fingertips. Available now for production environments, this latest version introduces a comprehensive suite of features designed to modernize workflows, streamline debugging, and enhance productivity. Among the headline additions are Bring Your Own Agent (BYOA) integration, stable Android XR emulators, an agent skills manager, AI-driven code reviews, and deep integration with LeakCanary for automated memory leak resolution. The release coincides with the ongoing Android Dev Summit, where engineering teams are showcasing live demonstrations and technical talks detailing how these new capabilities fit into modern development pipelines. As development workflows increasingly incorporate agentic tools to write and refactor code at unprecedented speeds, Android Studio Rabbit 2 aims to address both the acceleration of code creation and the emerging bottlenecks associated with reviewing and verifying that code. Read Also: Android Auto and Automotive OS Officially Graduate Games Category to General Availability Android Dev Summit Highlights Next-Generation AI Agent Integration, XR Emulators, and Performance Tools for Developers Bring Your Own Agent (BYOA): Integrate Your Preferred Coding Agents Flexibility has emerged as a central theme in modern software engineering, prompting the development team behind Android Studio to prioritize workflow customization. Recognizing that different projects and enterprises have unique requirements regarding cost efficiency, compliance, and capability, Android Studio Rabbit 2 introduces Bring Your Own Agent, or BYOA. This capability allows developers to access the most powerful coding agents alongside the robust suite of traditional Android development tools. Under this new framework, developers are no longer locked into a single ecosystem. Teams that prefer specific Agent Client Protocol (ACP) compliant agents can integrate them seamlessly. Whether an engineering organization requires the cost efficiency of a particular agent for a specific project or must adhere to strict enterprise agreements dictating the use of approved providers, the platform accommodates these needs. The integration supports a vast ecosystem of over 40 distinct coding agents. Notable options available out of the box include Claude Agent from Anthropic, Codex from OpenAI, Junie, and Google Antigravity. Crucially, these external agents are not isolated within a chat window; they possess direct access to vital Android Studio tools such as the Android Emulator, underlying build systems, and Logcat. Furthermore, they can leverage specialized Android skills to execute complex, multi-step tasks across the codebase. Getting started with BYOA involves navigating to the agent window, selecting a preferred agent or configuring options through the settings menu under AI agents, completing the authentication process, and submitting the first prompt. Build for Android XR with Emulators As the computing landscape expands beyond traditional mobile screens into spatial computing, wearable displays, and intelligent eyewear, developers require reliable tools to build immersive experiences. Addressing this demand, the Android XR Emulators for XR glasses, audio glasses, and display glasses have reached stable status within Android Studio Rabbit 2. These tools enable developers to build, iterate, and test sophisticated Android XR experiences directly from their laptops without requiring physical hardware for every stage of development. The updated display glasses emulator introduces powerful customization options, allowing developers to test applications within realistic environments by supplying their own 360-degree photos or a live video feed from a connected camera. This capability ensures that virtual environments can be simulated from anywhere in the world. Additionally, the release incorporates new hardware LED indicators designed to reflect microphone and camera states accurately. Improvements to the Pairing Wizard also streamline the setup process, enabling developers to pair a virtual phone with virtual audio and display glasses more quickly and reliably through the Device Manager. Discover, Install, and Manage Android Skills To maximize the effectiveness of coding agents, Android Studio Rabbit 2 introduces a dedicated skills manager. This tool allows developers to discover, install, and manage specialized Android skills directly within the integrated development environment, providing granular control over the procedural guidance and specialized workflows available to the AI. The environment comes pre-loaded with a rich set of Android skills and Firebase agent skills designed to handle common and complex architectural shifts. These include migrating legacy XML Views to modern Jetpack Compose, upgrading projects to Android Gradle Plugin (AGP) 9, implementing adaptive layouts tailored for foldables and tablets, and integrating complex libraries like CameraX alongside various Firebase services. Developers can access the skills editor through the agent panel, by typing the "@" symbol directly in the chat box, or via the settings menu, allowing them to invoke installed skills contextually during their coding sessions. Agentic Fixes with LeakCanary Debugging memory leaks has historically been a tedious and time-consuming aspect of mobile engineering. Android Studio Rabbit 2 transforms this diagnostic process by integrating LeakCanary directly with the built-in development agent. This native integration allows developers to diagnose and resolve memory issues without ever leaving the IDE. When a memory leak is detected in the Profiler, LeakCanary goes beyond traditional stack traces by providing detailed narrative explanations of the underlying causes. More importantly, the integrated agent suggests specific, actionable code changes designed to correct the root cause of the leak. This agentic approach bridges the gap between identifying performance bottlenecks and implementing solutions, significantly reducing the time required to maintain application stability. Gemini Enterprise and Guided Change Reviews For engineering teams operating under strict governance, data privacy mandates, and custom internal knowledge bases, Android Studio Rabbit 2 continues to offer robust support for Gemini Enterprise. Enterprise teams can connect the development environment directly to internal design systems, proprietary SDKs, and private repositories, ensuring that AI-generated responses remain highly relevant and secure. Complementing these enterprise capabilities is a new feature addressing the modern engineering bottleneck of code review. As agentic tools accelerate code generation and refactoring, verifying modifications across large codebases has become increasingly complex. The new AI Change Review feature meets developers wherever code changes occur within the IDE, accessible via visual entry points or directly from the agent chat window. The review process begins by identifying recommended fixes through an independent analysis of agent changes, pending commits, or git logs. Once these preliminary recommendations are addressed, developers can trigger a guided walkthrough. Rather than forcing reviewers to examine files in rigid alphabetical order, the guided walkthrough organizes and prioritizes related diff chunks across the project into logical themes, streamlining verification and ensuring high code quality standards are maintained across teams. Developers can begin exploring these features by downloading the latest stable version of Android Studio. The community can connect through official channels on LinkedIn, YouTube, and X, while ongoing discussions and demonstrations continue to stream as part of the Android Dev Summit. Post navigation Google Launches Android XR Platform, Stabilizes Core Tooling, and Welcomes XREAL AURA at Android Developer Summit