terraform

I took a look at your terraform skill and wanted to share some thoughts.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 74/100, solidly in C territory. This is based on Anthropic's skill best practices rubric. Your strongest area is Ease of Use (22/25)—the metadata and trigger patterns are really well done. The weakest spot is Progressive Disclosure Architecture (18/30)—everything's loaded at once, which burns tokens unnecessarily.

What's Working Well

• Metadata is chef's kiss. Your description clearly explains the READ-ONLY constraint, and your trigger patterns cover terraform commands, file types, and cloud provider prefixes (aws_, azurerm_, google_). That's thorough.
• Command safety is solid. The forbidden/allowed command lists are comprehensive and make the constraint crystal clear to users. This is exactly what a skill should do for sensitive domains.
• HCL patterns are practical. Your module structure, variable definitions, and lifecycle rules are real examples developers actually need.

The Big One: Token Efficiency (High Priority)

Right now, all 360 lines load every time someone uses your skill. That's wasteful. You need a references directory with on-demand-loaded files.

Here's the fix:

• Create references/aws-patterns.md (provider configs, common resources)
• Create references/lifecycle-rules.md (the detailed lifecycle examples)
• Create references/module-structure.md (complete module templates)

Keep only the essentials in SKILL.md—the critical safety restrictions and quick-start. The patterns section becomes a pointer: "See module structure reference for complete templates."

Impact: This single change gets you +8 points, pushing you to **82...