nda-triage
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of NDA documents provided by users or counterparties, which creates a potential surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the agent's context through the NDA document provided as a file path or direct text content.
- Boundary markers: The instructions lack explicit boundary markers or delimiters (such as XML tags or unique markers) to isolate the untrusted NDA content from the skill's logic.
- Capability inventory: The skill generates markdown reports and triage artifacts; it does not demonstrate use of high-risk capabilities like arbitrary command execution or network operations within the provided file.
- Sanitization: There are no instructions for sanitizing or escaping the content of the ingested NDA before processing.
Audit Metadata