total-wine-product-data

Pass

Audited by Gen Agent Trust Hub on Oct 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes external product data, including names, brands, and descriptions, retrieved from the Total Wine API through api.scavio.dev. This external content represents a potential surface for indirect instructions to influence agent behavior.
  • Ingestion points: Data returned from the /search and /product endpoints as described in the workflow and code examples in SKILL.md.
  • Boundary markers: The skill includes a "Guardrails" section in SKILL.md providing explicit instructions for the agent to stick to the returned API data and avoid fabrication, though no technical delimiters or sanitization logic are implemented in the code snippets.
  • Capability inventory: The skill performs POST requests to retrieve inventory and pricing information using standard HTTP libraries and the scavio SDK.
  • Sanitization: The documentation and examples do not specify procedures for escaping or filtering the strings returned from the API responses.
  • [EXTERNAL_DOWNLOADS]: The documentation provides instructions for installing the vendor's official libraries and tools.
  • Evidence: Mentions installation of the scavio package via pip and npm, and the @scavio/mcp-server via npx within SKILL.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 29, 2026, 03:25 PM