CtrlK
BlogDocsLog inGet started
Tessl Logo

adk-infra-expert

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.

76

1.29x

Quality

71%

Does it follow best practices?

Impact

84%

1.29x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./plugins/devops/jeremy-adk-terraform/skills/adk-infra-expert/SKILL.md
SKILL.md
Quality
Evals
Security

Evaluation results

86%

30%

Bootstrap ADK Agent Infrastructure

Terraform initialization structure and provider config

Criteria
Without context
With context

Top-level terraform dir

100%

100%

modules subdir

100%

100%

envs/dev subdir

0%

100%

envs/prod subdir

0%

100%

Terraform required_version

70%

100%

Google provider version

100%

100%

Google provider source

100%

100%

enable_code_execution variable

100%

100%

enable_code_execution default

0%

0%

enable_memory_bank variable

0%

100%

enable_memory_bank default

0%

0%

agent_name variable

100%

100%

Without context: $0.4735 · 2m 10s · 24 turns · 27 in / 8,042 out tokens

With context: $0.9476 · 3m 14s · 48 turns · 906 in / 11,151 out tokens

73%

ADK Agent IAM Module

Service account naming and IAM role assignment

Criteria
Without context
With context

Module directory exists

100%

100%

Service account resource

100%

100%

account_id naming pattern

0%

0%

display_name pattern

0%

0%

IAM binding resource

100%

100%

roles/aiplatform.user assignment

100%

100%

service account email output

100%

100%

agent_name variable in module

100%

100%

project_id variable in module

100%

100%

Module invocation example

100%

100%

Without context: $0.2234 · 1m 5s · 14 turns · 16 in / 3,424 out tokens

With context: $0.5459 · 2m 11s · 29 turns · 854 in / 7,152 out tokens

93%

26%

Production ADK Infrastructure for Regulated Industry

VPC-SC, code execution sandbox, and monitoring configuration

Criteria
Without context
With context

VPC-SC service perimeter

100%

100%

VPC with Private Service Connect

50%

100%

Code execution config

60%

100%

14-day code execution default

0%

100%

Memory Bank configuration

50%

100%

Cloud Monitoring dashboards

100%

100%

Cloud Monitoring alerts

100%

100%

VPC-SC data exfiltration rationale

100%

100%

Validation steps in checklist

100%

100%

Least privilege IAM

87%

100%

gcloud API enable command

0%

30%

Without context: $0.8923 · 5m 6s · 28 turns · 35 in / 18,294 out tokens

With context: $1.0219 · 4m 28s · 40 turns · 1,904 in / 16,392 out tokens

Repository
jeremylongshore/claude-code-plugins-plus-skills
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.