Practical overview
When it fits
Use it to scaffold or extend a Google Cloud-oriented agent project when its documented templates and infrastructure workflow fit the task. Consider the maintenance-mode notice before starting a new project.
Who it is for
- Developers building agent projects for Google Cloud.
- Teams choosing among documented patterns such as ReAct, retrieval-augmented generation, multi-agent communication or real-time multimodal agents.
- Developers who want project scaffolding for evaluation and deployment workflows around their agent code.
Documented uses
- Create an agent project from a reusable template.
- Explore agent patterns implemented with ADK, LangGraph or ADK for Java.
- Set up a project structure that includes evaluation, CI/CD, observability and deployment infrastructure.
- Add deployment and infrastructure scaffolding to an existing agent project.
Requirements & limits
- The project is in maintenance mode: only critical fixes are planned, with no new features, templates or deployment targets.
- New projects are directed to agents-cli; existing projects are directed to a migration guide.
- The documented minimum Python version is 3.10, and deployment requires Terraform; Google Cloud SDK and Make are also listed prerequisites.
- The repository is public and uses the Apache-2.0 license.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Create an agent project with uv
Scaffold a new project from the Agent Starter Pack CLI.
Before you start
- Python 3.10 or later.
- uv installed.
Steps
Run
uvx agent-starter-pack create.Explore and customize the generated agent project.
Expected outcome
A starter agent project with backend, frontend and deployment infrastructure to explore and customize.
Guide limitations
- The command scaffolds a project; deployment is a separate stage.
- Agent Starter Pack is in maintenance mode, with critical fixes only and no new features, templates or deployment targets.
Add starter-pack infrastructure to an existing agent
Add production-oriented deployment and infrastructure scaffolding to an existing agent project.
Before you start
- An existing agent project.
- Python 3.10 or later and the documented development prerequisites.
Steps
Open the existing project's root folder.
Run
uvx agent-starter-pack enhance.
Expected outcome
The README describes this command as adding deployment and infrastructure support to the existing project.
Guide limitations
- The command's documented purpose is enhancement; the README does not make this command itself a deployment step.
- The project is in maintenance mode.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 16 October 2026
- Original website title
- GoogleCloudPlatform/agent-starter-pack · Official homepage title · retained in its source language
Related options
agents-cli — The README identifies it as the successor with active development and recommends it for new projects; existing Agent Starter Pack projects are directed to migrate.

