dataset-management
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides standard code snippets for Dataiku dataset operations. All API calls are used according to official Dataiku documentation.
- [DATA_EXFILTRATION]: The skill includes code to read local files via the open() function for the purpose of uploading them to Dataiku datasets. This is an expected behavior for a dataset management tool and does not involve unauthorized data transmission.
- [PROMPT_INJECTION]: The skill possesses a data ingestion surface through local file reads for schema detection (SKILL.md, references/schema-operations.md). While boundary markers and sanitization are absent, the capabilities are limited to structured API calls (uploaded_add_file, autodetect_settings) and do not include arbitrary command execution.
Audit Metadata