patent-novelty-check

Pass

Audited by Gen Agent Trust Hub on Apr 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, hidden commands, or security risks were detected in the skill instructions or workflow. The skill uses tools within their intended scope for patent research.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes invention descriptions and prior art reports—which often contain external content from web searches—and incorporates this data into a prompt for a secondary AI review model via mcp__codex__codex.
  • Ingestion points: $ARGUMENTS (user-supplied invention description) and patent/PRIOR_ART_REPORT.md (content likely gathered from WebSearch).
  • Boundary markers: None present; external data is interpolated directly into the examiner verification prompt without delimiters or instructions to ignore embedded commands.
  • Capability inventory: The skill has access to Bash, Write, WebSearch, WebFetch, and the mcp__codex__codex tool.
  • Sanitization: No validation or sanitization of the input text is performed before it is passed to the reviewer model.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 19, 2026, 03:14 AM