fabricks-sql-analyzer

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: Fetches and executes the official Databricks CLI installation script from the vendor's public GitHub repository.
  • [COMMAND_EXECUTION]: Invokes shell commands for environment verification, Databricks authentication, and Git workflow management, including branch creation and commits.
  • [PROMPT_INJECTION]: The skill processes SQL files from the local repository which constitutes an indirect prompt injection surface.
  • Ingestion points: The scripts/sql_dependency_analyzer.py script scans and reads all .sql files in the repository.
  • Boundary markers: The agent is instructed to follow structured script output and apply a fixed set of transformations from a predefined catalog, using TODO comments for ambiguous cases.
  • Capability inventory: File writing (via Git and code editing), shell execution (bash, git, uv, databricks), and network access via the official Databricks API.
  • Sanitization: Employs the sqlglot library for structured parsing and sqlfmt for ensuring code consistency after automated edits.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 10:02 AM