The Agent Skills Directory

[SAFE] (SAFE): The skill's primary purpose is to enforce and document security best practices for AI deployment. It covers areas like container isolation, secrets management, and dependency auditing.
[DATA_EXPOSURE] (SAFE): No hardcoded credentials, sensitive file access, or unauthorized network exfiltration patterns were detected. The skill actually provides templates for using HashiCorp Vault to avoid hardcoding secrets.
[REMOTE_CODE_EXECUTION] (SAFE): The provided Python script (scripts/security-checklist.py) uses only standard Python libraries (json, typing, dataclasses, enum) and performs local logic without executing arbitrary commands or downloading remote code.
[PRIVILEGE_ESCALATION] (SAFE): The skill explicitly promotes the principle of least privilege, recommending non-root users, read-only filesystems, and dropping Linux capabilities in container configurations.

secure-deployment