multi-channel-selling
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill architecture involves ingesting data from external marketplace webhooks, which introduces a potential surface for indirect prompt injection.\n
- Ingestion points: Order data from Amazon and Walmart is received via HTTP webhooks, as described in the marketplace ingestion tasks.\n
- Boundary markers: The provided code and design tasks do not specify the use of data delimiters or instructions for the agent to ignore embedded commands within the marketplace payloads.\n
- Capability inventory: The skill is designed to perform database write operations (order creation, inventory updates) and outbound API calls to marketplace services.\n
- Sanitization: The instructions focus on structural normalization rather than explicit sanitization of field content to prevent adversarial data from influencing agent logic.
Audit Metadata