amazon-global-selling
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external marketplace data, including competitor listings and customer review insights, which represent potential surfaces for indirect prompt injection.
- Ingestion points: The 'Inputs and Collection' section (SKILL.md) specifies gathering competitor evidence and review insights from external marketplaces.
- Boundary markers: The workflow in 'Step 1' defines an evidence boundary with explicit labels (Confirmed, Directional, Unresolved) to categorize source data.
- Capability inventory: Capabilities across the skill include economic modeling, table generation, and creation of localization briefs based on the ingested data.
- Sanitization: The 'Localize the Offer' section (Step 6) includes instructions to preserve verified facts and prohibit unapproved claims, acting as a logical output filter.
- [EXTERNAL_DOWNLOADS]: The skill provides installation instructions utilizing a package from the vendor's own repository.
- Evidence: The installation section in SKILL.md refers to
npx skills add nexscope-ai/Amazon-Skills. This is a standard vendor-provided resource belonging to the skill author 'nexscope-ai'.
Audit Metadata