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.
51
57%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./plugins/devops/jeremy-adk-terraform/skills/adk-infra-expert/SKILL.mdQuality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong skill description that clearly defines its niche at the intersection of Vertex AI ADK and Terraform infrastructure provisioning. It excels in specificity by listing concrete resources provisioned, provides excellent trigger phrases, and explicitly addresses both what the skill does and when to use it. The only minor note is the slightly awkward opening 'Execute use when' which could be smoother, but it doesn't materially impact functionality.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions and components: 'Provisions Agent Engine runtime, 14-day code execution sandbox, Memory Bank, VPC Service Controls, IAM roles, and secure multi-agent infrastructure.' These are highly specific, named resources. | 3 / 3 |
Completeness | Clearly answers both 'what' (provisions Agent Engine runtime, sandbox, Memory Bank, VPC, IAM, multi-agent infrastructure) and 'when' (explicit 'Use when provisioning Vertex AI ADK infrastructure with Terraform' plus trigger phrases). Both are explicitly stated. | 3 / 3 |
Trigger Term Quality | Includes multiple natural trigger phrases users would say: 'deploy ADK terraform', 'agent engine infrastructure', 'provision ADK agent', 'vertex AI agent terraform', 'code execution sandbox terraform'. These cover realistic variations of how a user would request this task. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive niche combining Vertex AI ADK, Terraform, and specific infrastructure components like Agent Engine and code execution sandbox. Very unlikely to conflict with other skills due to the specificity of the domain. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
14%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is essentially a table of contents with no substantive content. It describes what should be done at a very high level but provides zero executable Terraform code, no concrete commands, and no validation steps for what is a complex, multi-step infrastructure provisioning workflow. All meaningful content is deferred to reference files that don't exist in the bundle, making the skill non-functional.
Suggestions
Add concrete, executable Terraform HCL blocks for at least the core resources (Agent Engine, IAM, VPC) directly in the SKILL.md or ensure the referenced bundle files actually exist.
Include explicit validation checkpoints with actual commands (e.g., 'terraform plan -out=tfplan', 'terraform validate', 'gcloud ai agent-engines describe') between provisioning steps.
Replace the abstract instruction list with a copy-paste-ready workflow including specific CLI commands like 'terraform init -backend-config=...', 'terraform apply', and verification commands.
Either provide the referenced bundle files (implementation.md, errors.md, examples.md) or inline the critical content so the skill is self-contained and functional.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The prerequisites section includes items Claude already knows or can infer (e.g., 'Terraform 1.0+ installed', 'gcloud CLI authenticated'). The overview restates the description. Some padding exists but it's not egregiously verbose. | 2 / 3 |
Actionability | No executable code, no concrete Terraform configurations, no specific commands. The instructions are entirely abstract descriptions ('Configure Variables', 'Provision VPC') without any actual HCL, CLI commands, or copy-paste-ready content. Everything actionable is deferred to reference files that don't exist in the bundle. | 1 / 3 |
Workflow Clarity | While steps are numbered, they lack any validation checkpoints, feedback loops, or error recovery guidance. For a destructive/provisioning operation like Terraform infrastructure deployment, there are no 'terraform plan' verification steps, no rollback instructions, and no explicit validation between steps. This should cap at 1 given the high-risk nature of infrastructure provisioning. | 1 / 3 |
Progressive Disclosure | References point to implementation.md, errors.md, and examples.md, but no bundle files are provided, meaning all substantive content is deferred to non-existent files. The SKILL.md itself is essentially an empty shell with no actionable content — the progressive disclosure pattern fails because the main file contains nothing useful on its own. | 1 / 3 |
Total | 5 / 12 Passed |
Validation
81%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 9 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
allowed_tools_field | 'allowed-tools' contains unusual tool name(s) | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 9 / 11 Passed | |
- Commit
69c73e9
- Reviewed
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.