signature-request
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection Surface: The skill is designed to ingest and process content from external documents such as PDF or DOCX files and cloud storage links. This data ingestion presents a surface where untrusted content could potentially contain instructions aimed at influencing the agent's behavior.
- Ingestion points: As described in Step 1, the skill accepts file uploads, URLs, and references to documents.
- Boundary markers: The skill instructions do not explicitly define delimiters or provide specific prompts for the agent to ignore instructions that might be embedded within the documents being processed.
- Capability inventory: The skill utilizes integrated tools for electronic signatures and document management to route files and send communications.
- Sanitization: There is no explicit content validation or sanitization step for the text extracted from documents prior to configuring the signature request.
Audit Metadata