THE DECODER·2 min read

Google's new Gemini API Agent Skill patches the knowledge gap AI models have with their own SDKs

Google's Gemini API enhances AI coding assistants.

Google has introduced an 'Agent Skill' for its Gemini API that resolves a critical issue faced by AI coding assistants: the lack of awareness about their own updates and best practices. This new skill provides coding agents with current information on models, SDKs, and sample code, leading to significant performance improvements. In tests involving 117 tasks, the Gemini 3.1 Pro Preview model's success rate surged from 28.2% to 96.6%, showcasing the skill's effectiveness, especially for newer models.

Key Takeaways

  • 1.

    The Gemini 3.1 Pro Preview model achieved a 96.6% success rate in coding tasks.

  • 2.

    Older models only saw minor improvements due to weaker reasoning capabilities.

  • 3.

    The skill is now available on GitHub for developers.

Get your personalized feed

Trace groups the biggest stories, videos, and discussions into one feed so you can stay current without scanning ten tabs.

Try Trace free
Google's new Gemini API Agent Skill patches the knowledge gap AI models have with their own SDKs | Trace