The Agent Skills Directory

[SAFE]: The skill explicitly instructs the agent to avoid hardcoding credentials, recommending the use of environment variables or GIS("home") for authentication, which aligns with security best practices.
[SAFE]: A robust safety mechanism is defined for destructive operations (e.g., truncate, delete, delete_features, overwrite). It requires the agent to identify the target, display its metadata (title, type, count), confirm the environment is non-production, and obtain explicit user confirmation before emitting any destructive code.
[SAFE]: The instructions emphasize checking runtime manifests and version compatibility for libraries like arcgis and arcpy rather than arbitrarily installing packages, which promotes environment stability.

python-notebook