ecommerce-patent-abstract-image-data

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes patent metadata and image paths retrieved from the PatSnap database via an external API call.
  • Ingestion points: The scripts/patent_abstract_image_data.py script fetches content from the ${NEXSCOPE_PROXY_BASE}/api/v1/tools/research/zhihuiya/abstractImage endpoint.
  • Boundary markers: There are no explicit instructions or delimiters defined in SKILL.md to warn the agent against executing instructions potentially embedded in the external API data.
  • Capability inventory: The skill uses urllib for network communications and writes JSON results to the local filesystem within the nexscope/ directory.
  • Sanitization: The script parses the response using standard JSON libraries but does not perform additional sanitization or filtering on the retrieved strings before they are processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 04:13 AM
Security Audit — agent-trust-hub — ecommerce-patent-abstract-image-data