data-pipelines
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill adheres to security best practices and does not contain malicious code, obfuscation, or unauthorized data exfiltration mechanisms.- [EXTERNAL_DOWNLOADS]: All referenced Python packages (e.g., dagster, dbt-core, boto3, dlt) are standard industry libraries sourced from the official Python Package Index (PyPI).- [COMMAND_EXECUTION]: The use of BashOperator and dbt CLI integration is restricted to standard orchestration tasks, such as triggering data transformations and builds.- [PROMPT_INJECTION]: While the skill defines patterns for ingesting external metadata (e.g., S3 object keys and GitHub issue data), this is a core requirement for pipeline orchestration and does not constitute a malicious injection vulnerability.- [CREDENTIALS_UNSAFE]: The skill explicitly instructs against hardcoding secrets, instead demonstrating the use of Airflow Connections, Dagster EnvVar resources, and cloud-native secrets managers.
Audit Metadata