The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill utilizes bash for routine environment checks like git logs and dependency listings. All commands are static and appropriate for the debugging context.
[DATA_EXPOSURE] (SAFE): Data collection is limited to error context and environment metadata. There are no attempts to access credentials, sensitive files, or exfiltrate data.
[INDIRECT_PROMPT_INJECTION] (SAFE): The skill processes external error messages and stack traces. While it lacks explicit separators, the structured methodology and lack of dynamic string execution in the shell mitigate the risk.

debugging