rowlab

Pass

Audited by Gen Agent Trust Hub on Nov 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from datasets that are subsequently used in prompts for AI models.
  • Ingestion points: Data enters the context via dataset_add, dataset_get, and sql_query in SKILL.md.
  • Boundary markers: The instructions recommend using run_preview to inspect the generated prompt, providing a point for manual verification, though no programmatic boundaries are enforced.
  • Capability inventory: The agent can perform SQL queries, trigger LLM-based labeling (run_start), and export data (dataset_export).
  • Sanitization: There are no specific instructions to sanitize or escape the content of the dataset rows before they are interpolated into the prompt templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 10, 2026, 12:23 PM