The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/check_ingestion_delay.py uses subprocess.run to interact with the platform's ae-cli tool. This is the primary function of the skill and is implemented using controlled argument lists rather than shell strings. The executable path is resolved via shutil.which and inputs are handled via argparse with specific character validation.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from SQL query results provided by external tools. The implementation includes extensive validation logic in scripts/validate_output.py and scripts/report_model.py to ensure schema compliance and data integrity. Furthermore, scripts/check_ingestion_delay.py explicitly sanitizes the event_name parameter to prevent SQL injection by blocking quotes and backslashes.

ae-data-delay-troubleshooter