The Agent Skills Directory

[SAFE]: The skill demonstrates no malicious intent or suspicious behavior. It focuses on internal workflow management and local data capture.
[DATA_EXFILTRATION]: No network operations, such as curl or wget, or access to sensitive credential files like .env or .ssh were detected.
[COMMAND_EXECUTION]: The skill does not execute arbitrary shell commands. It specifies that subcommands like append, list, and promote run inline without Bash or Task execution.
[PROMPT_INJECTION]: No attempts to bypass agent safety filters or override system instructions were found. The skill instructions are limited to task-specific logic and tool mapping.
[NO_CODE]: The skill does not include or download executable code or scripts.

gsd-note