dbt-unit-tests
Pass
Audited by Gen Agent Trust Hub on Apr 21, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: Uses the 'altimate-dbt' CLI tool to compile models, fetch column metadata, and execute unit tests. These commands are necessary for the skill's stated purpose and operate within the local project environment.
- [DATA_EXPOSURE]: The skill reads project files such as 'manifest.json' and dbt model SQL files to extract dependencies and transformation logic. This access is required for the tool to generate accurate tests and mock inputs.
- [INDIRECT_PROMPT_INJECTION]: The skill processes dbt model SQL and project manifests which are external data sources. While these could theoretically contain instructions to influence the agent, the risk is inherent to code analysis tools and managed through a structured workflow.
- Ingestion points: Reads dbt model SQL files and 'target/manifest.json' via the 'read' tool and 'dbt_unit_test_gen' function.
- Boundary markers: No explicit delimiters or instruction isolation warnings are used during file ingestion.
- Capability inventory: Includes bash command execution, and file system tools ('read', 'write', 'edit', 'glob').
- Sanitization: No specific sanitization or filtering of the ingested model code is described.
Audit Metadata