Index of every page on this site: llms.txt
All skills
Run it yourself

Execute use when provisioning Vertex AI ADK infrastructure with Terraform. Trigger with phrases like "deploy ADK terraform", "agent engine infrastructure", "provision ADK agent", "vertex AI agent terraform", or "code execution sandbox terraform". Provisions Agent Engine runtime, 14-day code execution sandbox, Memory Bank, VPC Service Controls, IAM roles, and secure multi-agent infrastructure.

Use this Skill: https://skilld.dev/gh/jeremylongshore/tons-of-skills-marketplace/adk-infra-expert

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencesexamples.md

≈1.2k tokens on demand. Your agent reads this file only when SKILL.md points to it.

ADK Terraform Examples

Example 1: Minimal Agent Engine Runtime

Deploy a basic ADK Agent Engine with default settings for development.

# variables.tf
variable "project_id" {
  description = "GCP project ID"
  type        = string
}

variable "region" {
  description = "GCP region for Agent Engine"
  type        = string
  default     = "us-central1"
}

# main.tf
module "adk_agent" {
  source     = "./modules/agent-engine"
  project_id = var.project_id
  region     = var.region

  agent_config = {
    display_name      = "dev-agent"
    model             = "gemini-2.0-flash"
    code_execution    = true
    memory_bank       = false
  }
}

output "agent_endpoint" {
  value = module.adk_agent.endpoint_uri
}

Validation commands:

terraform plan -var="project_id=my-project"
terraform apply -var="project_id=my-project" -auto-approve
gcloud ai agent-engines list --project=my-project --region=us-central1
curl -H "Authorization: Bearer $(gcloud auth print-access-token)" \
  "https://us-central1-aiplatform.googleapis.com/v1/projects/my-project/locations/us-central1/agentEngines"

Example 2: Production Agent with Memory Bank and VPC-SC

Full enterprise deployment with Memory Bank, Code Execution sandbox, and VPC Service Controls.

# terraform.tfvars
project_id           = "prod-ai-project"
region               = "us-central1"
enable_vpc_sc        = true
enable_memory_bank   = true
enable_code_execution = true
access_policy_id     = "123456789"

# main.tf
module "networking" {
  source     = "./modules/networking"
  project_id = var.project_id
  region     = var.region

  psc_config = {
    subnet_name  = "adk-psc-subnet"
    ip_cidr      = "10.0.1.0/24"
  }
}

module "iam" {
  source     = "./modules/iam"
  project_id = var.project_id

  service_accounts = {
    agent_runner = {
      display_name = "ADK Agent Runner"
      roles = [
        "roles/aiplatform.user",
        "roles/storage.objectViewer",
      ]
    }
    memory_writer = {
      display_name = "Memory Bank Writer"
      roles = [
        "roles/aiplatform.user",
        "roles/datastore.user",
      ]
    }
  }
}

module "adk_agent" {
  source     = "./modules/agent-engine"
  project_id = var.project_id
  region     = var.region

  agent_config = {
    display_name      = "prod-agent-v1"
    model             = "gemini-2.0-pro"
    code_execution    = var.enable_code_execution
    memory_bank       = var.enable_memory_bank
    service_account   = module.iam.service_accounts["agent_runner"].email
  }

  networking = {
    vpc_id     = module.networking.vpc_id
    subnet_id  = module.networking.psc_subnet_id
  }
}

module "vpc_sc" {
  count          = var.enable_vpc_sc ? 1 : 0
  source         = "./modules/vpc-service-controls"
  project_id     = var.project_id
  access_policy  = var.access_policy_id

  perimeter_config = {
    title    = "adk-perimeter"
    resources = ["projects/${var.project_id}"]
    restricted_services = [
      "aiplatform.googleapis.com",
      "storage.googleapis.com",
    ]
  }
}

Validation commands:

# Verify agent is reachable
gcloud ai agent-engines describe prod-agent-v1 \
  --project=prod-ai-project --region=us-central1

# Test Memory Bank connectivity
gcloud firestore databases list --project=prod-ai-project

# Verify VPC-SC perimeter
gcloud access-context-manager perimeters describe adk-perimeter \
  --policy=123456789

# Test agent invocation through PSC
curl -X POST \
  -H "Authorization: Bearer $(gcloud auth print-access-token)" \
  -H "Content-Type: application/json" \
  -d '{"input": "Hello agent"}' \
  "https://us-central1-aiplatform.googleapis.com/v1/projects/prod-ai-project/locations/us-central1/agentEngines/prod-agent-v1:query"

Example 3: Multi-Agent Architecture

Deploy multiple cooperating agents with shared Memory Bank.

locals {
  agents = {
    orchestrator = {
      model          = "gemini-2.0-pro"
      code_execution = false
      memory_bank    = true
    }
    researcher = {
      model          = "gemini-2.0-flash"
      code_execution = true
      memory_bank    = true
    }
    coder = {
      model          = "gemini-2.0-pro"
      code_execution = true
      memory_bank    = false
    }
  }
}

module "agents" {
  for_each   = local.agents
  source     = "./modules/agent-engine"
  project_id = var.project_id
  region     = var.region

  agent_config = {
    display_name      = each.key
    model             = each.value.model
    code_execution    = each.value.code_execution
    memory_bank       = each.value.memory_bank
    service_account   = module.iam.service_accounts["agent_runner"].email
  }
}

Tons of Skills by Intent Solutions | jeremylongshore.com

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No alerts2mo3 checks · Risk SAFE
  • Gen Agent Trust Hub2mo

    The skill provisions Google Cloud infrastructure using Terraform. It uses restricted shell commands and generates infrastructure code dynamically. While it identifies troubleshooting steps that involve temporary high-privilege roles, the skill's operations are otherwise standard and secure.

  • Socket2mo

    No alerts

  • Snyk2mo

    Risk: LOW · No issues

Signed by skilld at 820dfc8. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated last month
What it can do
Reads files Edits files Runs commands
version
2.23.0
author
Jeremy Longshore <jeremy@intentsolutions.io>
tags
[
  "devops",
  "deployment",
  "terraform",
  "iam"
]
All 7 allowed tools
ReadWriteEditGrepGlobBash(terraform:*)Bash(gcloud:*)
Other metadata
compatibility
Designed for Claude Code

README badge

README badge for jeremylongshore/tons-of-skills-marketplace/adk-infra-expert