Terraform Engineer

Use when implementing infrastructure as code with Terraform across AWS, Azure, or GCP. Invoke for module development, state management, provider configuration, multi-environment workflows, infrastructure testing.

This skill is a coherent, instruction-only Terraform specialist — it itself does not request credentials or install code, but following its guidance (configuring providers, remote backends, running terraform apply/tests) will require cloud credentials and access to state storage. Before using it: (1) never paste long-lived secrets or private keys into chat; prefer short-lived credentials, assume-role/session tokens, or managed identities; (2) review any generated Terraform code and backend configs before running terraform init/plan/apply; (3) run operations in an isolated/test environment first and avoid automatic 'apply' without human review; (4) ensure remote state access is tightly scoped (least privilege) and state buckets are encrypted and versioned; (5) if connecting CI, use secrets stored in your CI's secret manager rather than exposing them to the agent. If you want a higher-assurance review, provide samples of the Terraform code the skill would generate so you can inspect the exact changes before applying.

Type: OpenClaw Skill Name: terraform-engineer Version: 0.1.0 The OpenClaw AgentSkills bundle for 'terraform-engineer' is benign. All files contain documentation and code examples (HCL, Go, Rego, YAML) focused on Terraform best practices, including strong emphasis on security, state management, and testing. The SKILL.md provides clear, benign instructions for the AI agent, defining its role and workflow without any prompt injection attempts or malicious directives. References to sensitive files like `~/.aws/credentials` are within the context of standard Terraform configuration examples, not as instructions for exfiltration or unauthorized access. The skill explicitly warns against insecure practices like storing secrets in plain text.