optimizing-databricks-sql
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill follows a structured analysis workflow utilizing designated tools for SQL linting and validation, ensuring that optimization logic is grounded in platform-specific metadata.
- [SAFE]: Potentially hazardous or costly operations, such as executing ANALYZE TABLE or applying code changes to source files, are strictly gated behind explicit user confirmation prompts.
- [SAFE]: Data privacy is considered through recommendations to use non-invasive profiling algorithms when performing data comparisons, preventing the exposure of sensitive record values.
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the extraction and processing of SQL from user-provided notebooks, which acts as a primary ingestion point for untrusted data. The skill mitigates risks by employing specific boundary markers, such as mandatory confirmation gates for all file writes. Its capability inventory includes database execution and file modification, which are managed by routing all extracted content through a dedicated sanitization pass (altimate_core_check) to detect safety and PII issues.
Audit Metadata