cross-border-patent-risk-screen
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources which could potentially contain malicious instructions aimed at influencing the AI's risk assessment.
- Ingestion points: In Step 1, the skill processes product image URLs, e-commerce platform links (using
web_fetch), and text descriptions from the open web. - Boundary markers: The instructions do not specify any delimiters or safety guidelines for the agent to differentiate between the data being analyzed and potential instructions embedded within that data.
- Capability inventory: The skill utilizes web crawling (
web_fetch) and various specialized patent analysis MCP tools provided by the vendor. It does not possess high-risk capabilities such as arbitrary command execution, local file system modification, or sensitive credential access. - Sanitization: No explicit sanitization, validation, or filtering of the fetched external content is described before it is passed to the feature extraction and summarization components.
Audit Metadata