epo-patent-analyzer

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external patent documents which could contain embedded instructions. 1. Ingestion points: The skill reads application content using the Read tool. 2. Boundary markers: No explicit delimiters are specified for the untrusted input. 3. Capability inventory: The skill has access to Bash, Read, and Write tools. 4. Sanitization: No input filtering or escaping is described. This surface is inherent to the task of document analysis and is handled by the platform's standard safety guidelines.
  • [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute specialized compliance analyzers. This use of system commands is documented and strictly scoped to the skill's primary purpose of technical patent review.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 12:38 AM