<2-3 sentences summarising what the project addresses and solves, the target audience, and the benefit.> google/agents-cli provides a command-line interface and specialized skills designed to empower coding assistants, such as Gemini CLI or Claude Code, to build, evaluate, and deploy AI agents on Google Cloud. It aims to transform any coding assistant into an expert in managing the enterprise-grade agent lifecycle, simplifying complex cloud integrations and development workflows for engineers and researchers.

How It Works

This tool acts as an intermediary, equipping coding agents with the necessary commands and skills to interact with the Google Cloud agent stack and the Agent Development Kit (ADK). It abstracts away the intricacies of various CLIs and services, enabling agents to handle agent creation, scaling, governance, and optimization. The approach leverages ADK for agent framework capabilities and provides distinct skills for development workflows, Python API interaction, project scaffolding, evaluation methodologies, cloud deployment, and observability.

Quick Start & Requirements

• Installation: Install the CLI and skills via uvx google-agents-cli setup or add skills to a coding agent using npx skills add google/agents-cli.
• Prerequisites: Python 3.11+, uv, and Node.js are required.
• Documentation: Full documentation and tutorials are available.

Highlighted Details

• Extends coding assistants with comprehensive agent development lifecycle management.
• Facilitates building, scaling, governing, and optimizing enterprise-grade AI agents on Google Cloud.
• Integrates with Google Cloud services for deployment, CI/CD, observability, and Gemini Enterprise registration.
• Offers modular skills covering agent workflows, ADK Python API, project scaffolding, evaluation, and data ingestion.

Maintenance & Community

Feedback, bug reports, and feature requests are managed via GitHub Issues. Users can share projects or provide feedback directly via email at agents-cli@google.com. Contributions are primarily welcomed through issue-driven feedback, with the CLI distributed as a pre-built wheel for rapid iteration.

Licensing & Compatibility

The project is subject to Google Cloud Service Terms and specific "Pre-GA Offerings Terms." While local development is possible using an AI Studio API key, deployment and cloud-specific features necessitate a Google Cloud environment. Compatibility is noted with various coding agents.

Limitations & Caveats

This project is currently a "Preview" feature and operates under "Pre-GA Offerings Terms," which may entail limited support. Deployment and advanced cloud functionalities are dependent on a Google Cloud project.