Developing Genkit Python

This walks Claude through building AI apps with Genkit's Python SDK, which is apparently changing fast enough that the first thing it tells you is "don't trust your training data, check the docs." It covers the standard flow: Google AI as the default provider, the genkit start dev server, flows, tools, structured output, and FastAPI integration. The skill is basically a condensed reference layer with a heavy emphasis on common errors, which is honest. You'd reach for this when prototyping Genkit Python apps or debugging import and API issues. The examples are practical, the workflow is clear, and it assumes you know why you're here.