product-differentiation-ebay

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The SKILL.md file references an installation command that fetches the vendor's own packages via npx.
  • [PROMPT_INJECTION]: The skill processes external review data in scripts/analyzer.py, representing an indirect prompt injection surface. (1) Ingestion points: ReviewData lists in scripts/analyzer.py. (2) Boundary markers: None. (3) Capability inventory: None (no network, file-write, or shell execution found). (4) Sanitization: None. (5) Severity: SAFE.
  • [SAFE]: No patterns of data exfiltration, obfuscation, or persistence were found. The script uses only the Python standard library for its logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 11:37 PM