The Agent Skills Directory

[SAFE]: The skill enforces secure credential management by explicitly instructing the agent never to write credentials inline and to source them from the environment or user prompts instead.
[SAFE]: A robust safety protocol is implemented for destructive data operations like delete or truncate, requiring the agent to verify the target details and obtain explicit user confirmation before generating code.
[SAFE]: The instructions scope the agent's work to official ArcGIS libraries and runtimes, providing guidance that prevents the use of untrusted or non-standard dependencies.

python-notebook