Google has opened Android CLI for third-party AI agents - ForkLog: cryptocurrencies, AI, singularity, the future

# Google Opens Android CLI for Third-Party AI Agents

Google has released a stable version of Android CLI 1.0 and opened the tool for third-party AI agents. The company announced the launch at I/O 2026 conference.

Update components also include:

• Android Skills: a repository on GitHub with instructions in Markdown format. These are modular rule sets that help neural networks (Gemini, Claude Code, and others) better understand Android development specifics: from navigation to migrating to Compose;
• Knowledge Base: a specialized database accessible via the android docs command. It provides AI agents with fresh context from Android, Firebase, and Kotlin documentation, even if the model's training data is outdated.

Previously, Android CLI was mainly used with Gemini CLI and Google Antigravity. Now, the tool supports external agents like Claude Code or Codex. According to Google, this approach allows applications to be built three times faster.

Using Android CLI with Android Skills reduces token consumption of LLMs by approximately 70% compared to regular chatbots without platform context.

Integration with Android Studio

In version 1.0, the tool gained tight integration with the IDE. Through the new android studio command set, AI agents gained access to:

• error analysis and symbol declaration search;
• Compose Preview rendering;
• dependency version checks.

In fact, Android CLI has become a software layer on top of Android Studio, providing models with a "semantic" access to project structure and platform resources.

Google emphasized that development is not limited to its own models. Documentation explicitly mentions support for Claude and GPT. Android Studio itself now allows connecting any models, including local Gemma 4.

Android CLI has been integrated into the Antigravity 2.0 platform and Google AI Studio. The latter now features native development capabilities, including:

• automatic app code generation;
• built-in emulator;
• deployment to physical devices and publishing on Google Play (internal testing).

To automate complex tasks, the Journeys system has been implemented — it allows describing chains of actions (e.g., "go to cart and press pay") in simple words.

Despite expanding terminal capabilities, Google maintains Android Studio as the main hub for complex debugging and profiling, positioning CLI as a bridge between AI and professional development environment.

A pre-release version of Android CLI is already available for download. Developers can also use the command android sdk install for quick setup of necessary environment components.

Recall that in May, the company tightened policies on using search AI tools after a public experiment by BBC journalist Thomas Germain.