anomaly-detection

Pass

Audited by Gen Agent Trust Hub on Mar 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes Python scripts (scripts/enable_anomaly_detection.py and scripts/query_results.py) that perform administrative actions within the Databricks environment, such as creating or deleting schema-level monitors and executing SQL queries against system tables.\n- [PROMPT_INJECTION]: The scripts/query_results.py script exhibits a vulnerability to indirect prompt injection (SQL injection) by interpolating catalog and schema names directly into SQL strings. While common in administrative tools, it presents a risk if the parameters are sourced from untrusted data.\n
  • Ingestion points: catalog and schema parameters in scripts/query_results.py and scripts/enable_anomaly_detection.py.\n
  • Boundary markers: None identified.\n
  • Capability inventory: Ability to execute high-privilege SQL queries via spark.sql() and manage monitors via the Databricks SDK.\n
  • Sanitization: None; uses direct f-string interpolation for query construction.\n- [EXTERNAL_DOWNLOADS]: The skill references configuration and method signatures from Databricks' official SDK repository and documentation on ReadTheDocs. These are recognized as trusted vendor sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 8, 2026, 02:33 AM