amazon-analyse
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill ingests untrusted data from Amazon listings and reviews, which creates a surface for indirect prompt injection.
- Ingestion points: Data from Sorftime MCP API (SKILL.md).
- Boundary markers: None specified.
- Capability inventory: File writing via the Write tool (SKILL.md).
- Sanitization: None specified.
- [EXTERNAL_DOWNLOADS]: The skill makes network requests to mcp.sorftime.com to retrieve market data, which is its core purpose.
Audit Metadata