datalineage-bigquery-asset-impact-analysis

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the bq show command to retrieve JSON-formatted metadata about BigQuery datasets, specifically to discover the 'location' field. This is a legitimate use of standard Google Cloud management tools.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external metadata from the Google Cloud Data Lineage API and command-line outputs, which constitutes a potential attack surface for indirect prompt injection if the source data were maliciously crafted.
  • Ingestion points: Data returned by the DataLineageServer:search_lineage MCP tool and the output of the bq show command.
  • Boundary markers: The instructions do not specify explicit delimiters or 'ignore' instructions for the data being processed, though the data is expected to be structured metadata.
  • Capability inventory: The skill possesses the ability to execute shell commands (bq) and perform network requests to Google Cloud services.
  • Sanitization: There is no explicit mention of sanitizing or escaping asset names or transform process descriptions before they are included in the final Markdown report provided to the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 05:40 AM
Security Audit — agent-trust-hub — datalineage-bigquery-asset-impact-analysis