Does it support persistent state for code execution?

Yes, it configures the Vertex AI Code Execution Sandbox with a 14-day state TTL (Time-To-Live) by default.

How does this skill handle security?

It implements enterprise-grade security by setting up VPC Service Control perimeters and applying least-privilege IAM roles specifically for native agent identities.

What specific GCP services does this skill provision?

It provisions Vertex AI Agent Engine runtimes, Code Execution Sandboxes, Memory Banks, VPC Service Controls (VPC-SC), and required IAM service accounts and roles.

Is this skill compatible with standard Terraform workflows?

Yes, it generates standard HCL (HashiCorp Configuration Language) modules that can be integrated into existing Terraform CI/CD pipelines.

Vertex AI Agent Infrastructure Expert

Name: Vertex AI Agent Infrastructure Expert
Author: jeremylongshore

byjeremylongshore

•

884

•

Cloud Infrastructure

Provisions enterprise-grade Vertex AI Agent Engine infrastructure using Terraform and secure VPC configurations.

This skill automates the deployment of production-ready Vertex AI ADK environments, handling complex configurations like Agent Engine runtimes, secure Code Execution Sandboxes with 14-day state retention, and integrated Memory Banks. It ensures security and compliance by automatically generating VPC Service Controls (VPC-SC) perimeters and applying least-privilege IAM roles for native agent identities, making it an essential tool for developers building secure, multi-agent AI applications on Google Cloud Platform.

Key Features

01Secure Code Execution Sandbox setup with 14-day state persistence

02Least-privilege IAM identity configuration for AI agents

03884 GitHub stars

04Automated Agent Engine runtime provisioning with Gemini integration

05Enterprise VPC Service Controls (VPC-SC) perimeter management

06Native Memory Bank and tool-calling infrastructure deployment

Use Cases

01Standardizing Infrastructure as Code (IaC) for Vertex AI ADK projects

02Deploying production-ready AI agents with persistent code execution capabilities

03Implementing VPC Service Controls for sensitive generative AI workloads

Key Features

01Secure Code Execution Sandbox setup with 14-day state persistence

02Least-privilege IAM identity configuration for AI agents

03884 GitHub stars

04Automated Agent Engine runtime provisioning with Gemini integration

05Enterprise VPC Service Controls (VPC-SC) perimeter management

06Native Memory Bank and tool-calling infrastructure deployment

Use Cases

01Standardizing Infrastructure as Code (IaC) for Vertex AI ADK projects

02Deploying production-ready AI agents with persistent code execution capabilities

03Implementing VPC Service Controls for sensitive generative AI workloads