🚀 Getting Started
This guide quickly walks you through setting up your first agent project.
Want zero setup? 👉 Try in Firebase Studio or in Cloud Shell
Prerequisites
Python 3.10+ | Google Cloud SDK Install Guide | Terraform Install Guide | uv
(Optional, Recommended) Install Guide
1. Create Your Agent Project
You can use the pip
workflow for a traditional setup, or uvx
to create a project in a single command without a permanent install. Choose your preferred method below.
# 1. Create and activate a virtual environment
python -m venv .venv
source .venv/bin/activate
# 2. Install the package
pip install agent-starter-pack
# 3. Run the create command
agent-starter-pack create my-awesome-agent
# This single command downloads and runs the latest version
uvx agent-starter-pack create my-awesome-agent
No matter which method you choose, the create
command will:
- Let you choose an agent template (e.g.,
adk_base
,agentic_rag
). - Let you select a deployment target (e.g.,
cloud_run
,agent_engine
). - Generate a complete project structure (backend, optional frontend, deployment infra).
Examples:
# You can also pass flags to skip the prompts
agent-starter-pack create my-adk-agent -a adk_base -d agent_engine
2. Explore and Run Locally
Now, navigate into your new project and run its setup commands.
cd my-awesome-agent && make install && make playground
Inside your new project directory (my-awesome-agent
), you'll find:
app/
: Backend agent code.deployment/
: Terraform infrastructure code.tests/
: Unit and integration tests for your agent.notebooks/
: Jupyter notebooks for getting started with evaluation.frontend/
: (If applicable) Web UI for interacting with your agent.README.md
: Project-specific instructions for running locally and deploying.
➡️ Follow the instructions in your new project's README.md
to run it locally.
Next Steps
You're ready to go! See the Development Guide for detailed instructions on extending, customizing and deploying your agent.
- Add Data (RAG): Configure Data Ingestion for knowledge-based agents.
- Monitor Performance: Explore Observability features for production monitoring.
- Deploy to Production: Follow the Deployment Guide to deploy your agent to Google Cloud.