The Agent Skills Directory

[SAFE]: The skill is authored by 'databricks-solutions', a trusted vendor, and adheres to official Databricks documentation and patterns for data quality management.\n- [SAFE]: No prompt injection, behavioral override, or system prompt extraction patterns were detected in the skill markdown or code examples.\n- [SAFE]: The skill implements dynamic SQL generation to load data quality rules from a Delta table. This is an intended architectural feature for Spark Declarative Pipelines and does not represent a command execution risk in this context, as it relies on internal environment configuration and governed metadata tables.\n- [PROMPT_INJECTION]: The skill was evaluated for indirect injection vulnerabilities through its rule-loading mechanism. Ingestion point: dq_rules_loader.py reads data from the {catalog}.{schema}.dq_rules table. Boundary markers: None. Capability inventory: Spark SQL filtering and expectation decorators. Sanitization: None. The surface area is considered safe for its intended use as a pipeline configuration pattern.

dlt-expectations-patterns