LogoAIAny
Icon for item

Agent Starter Pack

Agent Starter Pack is a Python package providing production-ready templates for GenAI agents on Google Cloud. Focus on your agent logic—the starter pack provides everything else: infrastructure, CI/CD, observability, and security. Includes pre-built templates like ReAct, RAG, multi-agent, and more.

Introduction

Agent Starter Pack

Version Docs

A Python package that provides production-ready templates for GenAI agents on Google Cloud. It accelerates agent development by handling all non-core aspects like infrastructure, CI/CD, observability, and security, allowing developers to focus purely on agent logic.

Key Features
  • Quick Launch: Get started in 1 minute with uvx agent-starter-pack create my-awesome-agent or pip installation.
  • Pre-built Agents: Templates include adk_base (ReAct using Google's Agent Development Kit), adk_a2a_base (with Agent2Agent Protocol), agentic_rag (RAG with Vertex AI Search/Vector Search), langgraph_base (LangGraph ReAct), adk_live (real-time multimodal RAG with Gemini).
  • Deployment: Production infra on Cloud Run or Agent Engine, with monitoring and observability.
  • CI/CD: One-command setup for Google Cloud Build or GitHub Actions.
  • Enhancements: Add data pipelines for RAG, remote templates, Gemini CLI integration for guidance.
  • Experimentation: Vertex AI evaluation and interactive playground.
Get Started
uvx agent-starter-pack create my-awesome-agent

Or enhance existing agents: uvx agent-starter-pack enhance.

Architecture

Covers prototyping, evaluation, deployment, and monitoring. Integrates with Google Cloud SDK, Terraform, etc.

Community & Resources

Contributions welcome! See CONTRIBUTING.md.

Information

  • Websitegithub.com
  • AuthorsGoogle Cloud Platform
  • Published date2025/02/27

More Items