The Agent Skills Directory

[SAFE]: The skill's primary purpose is to establish safety guardrails (Careful, Freeze, and Guard modes) to protect against accidental file deletion or database modifications.
[COMMAND_EXECUTION]: While the skill mentions high-risk commands such as rm -rf and DROP TABLE, it does so in a defensive context, instructing the agent to warn the user and require explicit confirmation before execution.
[PROMPT_INJECTION]: No evidence of prompt injection, role-play bypasses, or attempts to override the underlying system's safety guidelines were detected.
[DATA_EXFILTRATION]: No network operations, credential usage, or sensitive data access patterns are present in the skill.

gstack-safety