The Agent Skills Directory

[SAFE]: The skill provides standard validation workflows for common Infrastructure-as-Code (IaC) tools such as Terraform, Ansible, Docker, and Kubernetes using established CLI tools.
[SAFE]: Instructions promote security best practices, including secret management, resource limits, and non-root container execution, and provide clear red flags for dangerous configurations.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of IaC configuration files and possesses high-privilege capabilities such as executing shell commands (e.g., terraform, docker, kubectl). While no specific exploits are present, the processing of external configurations that lack explicit boundary markers represents a potential attack surface.
Ingestion points: SKILL.md (activation triggers for *.tf, Dockerfile, playbook*.yml, etc.)
Boundary markers: None specified for the content of the processed files.
Capability inventory: Shell command execution across multiple scripts (e.g., terraform plan, docker build, kubectl --dry-run).
Sanitization: No explicit sanitization or filtering of external configuration content is mentioned prior to execution.

infrastructure-validation