The Agent Skills Directory

[SAFE]: The skill implements a security auditing framework for Terraform codebases, focusing on infrastructure best practices like Least Privilege, Network Isolation, and Data Encryption.
[SAFE]: The allowed_tools are restricted to Read, Glob, and Grep, which are passive, read-only operations used for scanning local files without risk of system modification or data exfiltration.
[SAFE]: No malicious patterns such as prompt injection, obfuscation, or remote code execution were detected. The examples of 'insecure' code are clearly labeled for instructional purposes and are not executable instructions.
[SAFE]: The skill promotes secure secrets management by instructing the agent to check for sensitive variables and the presence of secret-scanning hooks.

terraform-security-audit