analysis-artifacts
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from BigQuery, which represents a surface for indirect prompt injection. 1. Ingestion points: Data retrieved via BigQuery queries is processed to generate summaries and visualizations (SKILL.md). 2. Boundary markers: No explicit delimiters or boundary instructions are defined to separate untrusted query results from agent instructions. 3. Capability inventory: The agent can write files (SQL, Python, CSV, Markdown) and is instructed to generate Python scripts for execution (SKILL.md). 4. Sanitization: The workflow does not include steps for sanitizing or validating external data before interpolation into reports or script logic.
- [DYNAMIC_EXECUTION]: The skill requires the agent to generate and save Python scripts for data visualization purposes. Finding: Step 4 of the workflow instructs the agent to 'write a Python script to generate it and save both the script and the output image' (SKILL.md).
Audit Metadata