jp-patent-claims-review
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and parses untrusted data from user-provided PDF, Word, and text documents to perform its analysis.
- Ingestion points: User-uploaded patent application files (SKILL.md, Step 1).
- Boundary markers: The instructions do not define explicit delimiters or 'ignore' instructions for the parsed document content to prevent the agent from obeying instructions embedded within the files.
- Capability inventory: The skill is capable of writing HTML files to the workspace (@skill_workspace/report.html) and invoking secondary tools (novelty-check and non-obviousness-check skill) based on the parsed content.
- Sanitization: There is no evidence of sanitization, escaping, or validation of the text extracted from the uploaded documents before it is processed by the model or interpolated into the report.
Audit Metadata