korean-patent-search

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes and presents untrusted data (patent titles and abstracts) from an external source to the agent.
  • Ingestion points: Data enters the context through the XML response from the KIPRIS Plus API, specifically fields like inventionTitle and astrtCont (abstract) processed in scripts/patent_search.py.
  • Boundary markers: The skill does not implement delimiters or explicit 'ignore embedded instructions' warnings when presenting the retrieved patent data to the agent.
  • Capability inventory: The associated script scripts/patent_search.py is limited to making network requests to the official KIPRIS domain and parsing XML; it does not possess capabilities for file system modification or arbitrary command execution.
  • Sanitization: No content filtering or instruction-stripping is performed on the retrieved patent metadata before it is interpolated into the agent's output.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 01:00 AM
Security Audit — agent-trust-hub — korean-patent-search