figure-description

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a utility for patent documentation and does not contain malicious code or exfiltration patterns.
  • [DATA_EXFILTRATION]: The skill reads intellectual property documents like INVENTION_DISCLOSURE.md and CLAIMS.md, but its output is strictly directed to the local patent/figures/ directory. No network exfiltration instructions were found.
  • [PROMPT_INJECTION]: The skill has an ingestion surface for untrusted data. Ingestion points: User-provided images (Step 2) and markdown files (Step 1). Boundary markers: absent. Capability inventory: Bash, Read, Write, Edit, WebSearch, WebFetch. Sanitization: absent. While the surface for indirect prompt injection exists, there is no evidence of exploitation or malicious intent in the provided instructions.
  • [COMMAND_EXECUTION]: The Bash(*) tool is listed in the configuration but is not used in any of the workflow steps or instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 12:29 PM