omni-to-dbt-metricflow

Pass

Audited by Gen Agent Trust Hub on Nov 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted metadata (labels, descriptions, AI context) from an Omni instance. Ingestion points: omni models yaml-get output from source views and topics. Boundary markers: Explicit instructions to treat fetched data as translation input and stop if instructions are detected. Capability inventory: Package installation (pip install), configuration writes (omni models yaml-create), and database queries (omni query run). Sanitization: Mandatory user reporting and halting for suspicious fetched content.
  • [EXTERNAL_DOWNLOADS]: The skill suggests the installation of the dbt-metricflow tool and database-specific adapters via pip. These are well-known, reputable packages from the dbt Labs ecosystem.
  • [COMMAND_EXECUTION]: The workflow relies on standard developer CLI tools, including omni, dbt, mf, rg, tail, and xxd, to manage migration tasks and perform integrity checks. All execution is scoped to the migration process.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 10, 2026, 02:50 PM