The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/mx_pipeline.py uses importlib.import_module() within the _toolkit function. This is used to load internal toolkit modules (mx_audit, mx_validate) from the trusted sibling skill scenario-maya-expert. The module names are hardcoded in calls throughout the script, posing no risk of arbitrary code injection.
[COMMAND_EXECUTION]: The skill manages batch processing by spawning mayapy (Maya's standalone Python interpreter) processes using the mx_run utility. This behavior is the intended primary function of a pipeline scripting skill and follows best practices for process isolation to prevent scene crashes from halting a full batch run.
[DATA_EXPOSURE]: The preflight function perform checks for local network resolution of localhost and the machine's host name using socket.gethostbyname(). These operations are limited to local connectivity checks aimed at preventing known performance hangs in Maya on macOS and do not involve data exfiltration.
[REMOTE_CODE_EXECUTION]: While the skill instructions mention installing sibling skills via npx skills add, this is documented as a manual user action for environment setup rather than an automated or silent execution of remote scripts.
[SAFE]: The skill demonstrates high security maturity by explicitly disabling Maya script nodes during batch operations and bridge sessions, which mitigates risks associated with malicious embedded scene scripts.

scenario-maya-pipeline-scripting