tabular-review-lawvable
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, untrusted document content which serves as an attack surface for indirect prompt injection.
- Ingestion points: Documents are read using the
pdfanddocxskills inSKILL.md(Step 3). - Boundary markers: The agent prompt template lacks clear delimiters (e.g., XML tags or unique separators) or explicit instructions to ignore commands embedded within the text of the documents being processed.
- Capability inventory: The skill has capabilities to write to the local file system (creating Excel files via the
xlsxskill) and launch further sub-tasks, which could be leveraged if an injection is successful. - Sanitization: No evidence of sanitization, escaping, or validation of the extracted text was found before it is interpolated into the Excel output or processed by subsequent agent steps.
Audit Metadata