patent-examiner
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from patent applications and search results, which could contain instructions designed to alter the agent's behavior.
- Ingestion points: Patent application specifications, claims, abstracts, figures, and prior art disclosures (SKILL.md).
- Boundary markers: Absent; the instructions do not specify any delimiters or safety markers to isolate untrusted data.
- Capability inventory: The skill writes analysis results to the local filesystem at the
patents/analysis/directory (SKILL.md). - Sanitization: Absent; there is no logic for validating or sanitizing the content of the ingested materials.
- [EXTERNAL_DOWNLOADS]: The skill directs the agent to retrieve information from well-known services such as the USPTO, Google Patents, and Google Scholar to facilitate the prior art search.
Audit Metadata