Skip to content

Genkit Python Developer

Helps an agent build Python AI apps with Genkit: generation, tools, flows, and multi-turn agents.

Repository
google/skills
Stars
21,782
Licence
Apache-2.0
Files
15 files

What it does

It teaches an AI agent the conventions of Genkit, Google's SDK for building AI features in Python. Once installed, the agent knows how to write generate, streaming, tool, flow and multi-turn agent code, which imports to use, and how to run apps under the Genkit CLI so traces of model calls and tool use can be inspected. It also points the agent to reference notes on setup, dotprompt files, evals and common errors.

Great for

  • Python developers starting a new Genkit app that calls Gemini models
  • Developers building chatbots or assistants who need multi-turn agent patterns
  • Teams debugging Genkit errors and wanting trace-based verification

Skip it if

  • Non-Python projects, since the skill is specific to the Python SDK
  • Simple one-off API calls where the Genkit framework adds unnecessary complexity
  • Production deployments without review: the Agents API is marked Beta and some referenced files (such as setup.md and fastapi.md) were not included in the files reviewed

Try asking

  • “Build a multi-turn support chatbot in Python with Genkit agents and session history”
  • “Add a tool-calling flow that looks up order data and returns structured output”
  • “Debug why my Genkit app's tools never run by capturing traces with genkit start”
  • “Set up an evaluation dataset and evaluators for my Genkit flow”
Step-by-step guide: how to use Genkit Python DeveloperInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 15 original files, the Apache-2.0 licence and an attribution note.

    developing-genkit-python.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip developing-genkit-python.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip developing-genkit-python.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Use genkit start -- uv run src/main.py rather than uv run alone when you need traces to verify tool calls
  • Use the --non-interactive flag and flow:run for scripted checks, since genkit start runs until stopped with Ctrl+C
  • Keep GEMINI_API_KEY in an environment variable or secret store, not in source code

What's inside

  • references/agents-artifacts.md
  • references/agents-background.md
  • references/agents-branching.md
  • references/agents-custom.md
  • references/agents-http.md
  • references/agents-human-in-the-loop.md
  • references/agents-sessions.md
  • references/agents-state.md
  • references/agents.md
  • references/common-errors.md
  • references/dev-workflow.md
  • references/dotprompt.md
  • references/evals.md
  • references/examples.md
  • SKILL.md

Redistributed unmodified from google/skills under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.