dbt-transformation-patterns
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill content is purely educational, providing production-ready patterns for data transformation within the dbt ecosystem. No malicious command execution or unauthorized data access patterns were identified.
- [SAFE]: External links in the resources section point to official documentation, the dbt package hub, and community forums. These are well-known and trusted services in the data engineering industry.
- [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for processing external data, which inherently introduces a surface for indirect injection if that data is later used by an LLM, but the skill itself implements best practices for isolation.
- Ingestion points: Data enters the transformation DAG through source models defined in
models/staging/using the{{ source() }}macro. - Boundary markers: Data processing logic is strictly bounded by SQL DDL/DML and Jinja2 templating syntax.
- Capability inventory: The transformation workflow is managed by standard
dbtCLI commands (run,test,build). - Sanitization: Staging patterns demonstrate attribute cleaning, such as using
lower()on email fields and converting currency units.
Audit Metadata