skills/archetypeai/agent-skills/atai-operational-state-monitoring-agent-data-prep/Gen Agent Trust Hub
atai-operational-state-monitoring-agent-data-prep
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external sensor recordings (CSV/Parquet) for processing. While this provides a data ingestion surface, the risk is mitigated by the skill's primary focus on mathematical transformation and the inclusion of a dedicated validation script (preflight_roles.py) that verifies data against strict physical and structural rules. \n- [COMMAND_EXECUTION]: The skill employs standard Python file system utilities and parallel processing (ProcessPoolExecutor) to manage data workflows. These actions are limited to local file manipulation within user-specified directories and are consistent with the skill's data preparation functionality.
Audit Metadata