terraform-skill
by antonbabenkoterraform-skill is a diagnose-first skill for Terraform and OpenTofu work. Use it to review, debug, or plan changes across modules, tests, CI, scans, and state operations with version-aware guidance. It helps reduce identity churn, secrets exposure, blast radius, CI drift, and state corruption.
This skill scores 84/100, which means it is a solid directory listing for users who work with Terraform/OpenTofu and want diagnose-first guidance instead of generic prompting. The repo shows real workflow depth: a clear trigger in SKILL.md, an explicit response contract, version-aware guardrails, and seven detailed reference files covering CI/CD, code patterns, module design, security/compliance, state management, testing, and quick commands.
- Strong triggerability: the frontmatter says to use it for writing, reviewing, or debugging Terraform/OpenTofu modules, tests, CI, scans, and state ops, with failure modes named up front.
- Good operational clarity: SKILL.md defines a response contract with assumptions, risk category, remediation tradeoffs, validation plan, and rollback notes, which reduces agent guesswork.
- High agent leverage: the repository includes multiple focused references and concrete command/code examples for common Terraform workflows like validation, testing, security scanning, and state management.
- No install command is provided in SKILL.md, so users may need to infer activation/setup steps from the skill file structure.
- The skill is broad and opinionated; teams with very narrow Terraform conventions may need to adapt the guidance, especially around version-specific behavior and workflow execution paths.
Overview of terraform-skill skill
What terraform-skill is for
terraform-skill is a diagnose-first skill for Terraform and OpenTofu work. It helps you review, debug, or plan changes for modules, tests, CI pipelines, scans, and state operations with version-aware guidance.
Who should install it
Use the terraform-skill skill if you manage infrastructure as code and need better answers than a generic prompt can give. It is especially useful for platform engineers, DevOps teams, and anyone doing terraform-skill for Deployment, module refactors, or state-safe edits.
Why it stands out
The main value is decision quality: it forces context capture, identifies the risk class first, and ties recommendations to validation and rollback. That makes it stronger for real Terraform work where the wrong fix can create identity churn, secrets exposure, blast radius, CI drift, or state corruption.
When it is a good fit
Install terraform-skill if your work depends on knowing the runtime (terraform or tofu), exact version, backend, execution path, and environment criticality before changing code. It is a strong fit when you want the skill to explain tradeoffs, not just generate HCL.
How to Use terraform-skill skill
Install and load the right files
Install with:
npx skills add antonbabenko/terraform-skill --skill terraform-skill
For the best terraform-skill install outcome, read SKILL.md first, then inspect:
references/quick-reference.mdreferences/state-management.mdreferences/security-compliance.mdreferences/code-patterns.mdreferences/module-patterns.mdreferences/ci-cd-workflows.mdreferences/testing-frameworks.md
Give the skill the inputs it needs
A strong terraform-skill usage prompt names the toolchain and the decision boundary. Include:
- Terraform or OpenTofu version
- Provider names and major versions
- Backend type and state location
- Local, CI, Cloud, or Atlantis execution path
- Whether the environment is dev, staging, or production
- The exact goal: debug, refactor, validate, test, or release
Example prompt shape:
“Using terraform-skill, review this module for a production deployment. I’m on Terraform 1.10, AWS provider 5.x, remote S3 state, GitHub Actions CI. I need the safest fix for a failing plan after a for_each refactor.”
Use a workflow, not a one-off question
The best terraform-skill guide flow is:
- State the runtime and risk.
- Paste the smallest relevant HCL, plan error, or test failure.
- Ask for diagnosis first, not code generation.
- Request validation commands and rollback notes.
- Apply only after reviewing the plan artifact.
This skill is designed to answer with assumptions, risk category, remediation, validation, and rollback. That structure is what makes it useful for deployment decisions.
What to ask for first
If you are unsure where to start, ask for:
- a failure-mode diagnosis
- a safe refactor plan
- a CI/test matrix for your repo
- state migration steps
- security or compliance review of the change
terraform-skill skill FAQ
Is terraform-skill only for writing Terraform?
No. It also helps with OpenTofu, CI pipelines, testing, state recovery, and security review. The terraform-skill skill is broader than module authoring, but still focused on infrastructure-as-code work.
How is it different from a normal prompt?
A normal prompt may produce plausible HCL. terraform-skill is better when you need version-aware guidance, explicit assumptions, and deployment-safe validation steps. It is aimed at reducing risky edits, not just increasing output volume.
Is terraform-skill beginner-friendly?
Yes, if you can provide basic context. Beginners get the most value when they share the error, target environment, and intended change. The skill can then explain the likely failure mode and what to verify before apply.
When should I not use it?
Do not use it as a substitute for live cloud access, provider docs, or approval workflows. If you do not know the runtime, backend, or environment, the answer will be less precise. It is also not the right tool for unrelated scripting or generic DevOps questions.
How to Improve terraform-skill skill
Give tighter source material
Better inputs produce better fixes. Instead of “my plan fails,” provide the relevant resource block, variable values, and the exact error. For terraform-skill in production work, include the plan snippet or state action that changed.
Specify the risk you care about most
The skill is strongest when you name the primary risk: identity churn, secret exposure, blast radius, CI drift, compliance gaps, or state corruption. That helps it choose the right remediation and the right validation commands instead of giving a vague fix.
Ask for the smallest safe change
If you want a reviewable result, ask for the least disruptive option first. For example: “Fix this without replacing resources,” or “preserve state addresses during refactor.” That pushes the skill toward state-safe patterns rather than fresh rebuilds.
Iterate with validation evidence
After the first answer, feed back fmt, validate, plan, terraform test, or policy output. That is the fastest way to refine a terraform-skill guide response because the skill can narrow from diagnosis to a specific remediation and rollback path.