Overview
google-genai-skills is a repository of reusable agentic skills designed to expand how AI assistants work with Google AI frameworks and models. Its scope includes development workflows involving Gemini CLI and Antigravity IDE, with compatibility also described for Claude Code.
The collection packages focused guidance for particular SDKs and model-building tasks rather than presenting a single general-purpose assistant. Each listed skill addresses a defined part of the Google generative AI development ecosystem.
Included Skills
The repository identifies three implemented skills:
- google-adk-python covers the Python SDK for Google Agent Development Kit.
- google-genai-sdk-python provides best-practice guidance for the Google GenAI Python SDK.
- veo-build focuses on building with Google DeepMind Veo.
Together, these skills span agent development, Python SDK usage, and Veo-oriented work. Their implemented status distinguishes the currently available collection from roadmap items planned for future releases.
Installation and Requirements
The project distributes releases as downloadable archives. The documented process consists of selecting the latest release, downloading the appropriate compressed file, extracting it locally, and locating the main application file in the extracted folder. Individual skill guidance is kept in the repository’s documentation folder.
Supported environments are described as Windows 10 or later, macOS 10.15 or later, and modern Linux distributions. The stated baseline is 4 GB of memory with 8 GB recommended, alongside an Intel i5 or equivalent processor, with an Intel i7 or better recommended.
Project and Licensing
google-genai-skills is published under the MIT License and accepts community contributions through the repository workflow. Its roadmap includes skills for additional Google AI tools, expanded documentation and usage guides, and improvements informed by user feedback.