avada-commerce

Warn

Audited by Snyk on Nov 15, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (medium risk: 0.30). The required workflow runs arbitrary AVADA Commerce actions (Orders, Customers, Products) through Membrane, and storefront customers can write free text such as order notes, contact details, and reviews that these records return to the agent without it selecting a specific item.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

  • Direct money access detected (medium risk: 0.30). No payment, refund, transfer, charge, or trade is documented. The skill describes generic AVADA Commerce data management and workflow automation, lists "Orders" among the platform areas (SKILL.md:24), and offers a generic API proxy that injects stored credentials (SKILL.md:130). These are capability hints, and the tiering rules say they do not by themselves establish a payment operation. Under the recall-first override, I treat order and sales management on a commerce platform as plausibly including direct financial actions, such as refunds or order-related charges, that the agent could reach through membrane action run or the proxy. This is an inference, not a documented route. If such a direct operation exists, the Membrane connection requires an authenticated service session (SKILL.md:41-47, 59-66), which is service-enforced access control, so the tier would be Medium rather than High. Indirect operations also reach Medium. Medium is therefore the recall-first answer, with the inferred nature of the route stated plainly.

Issues (2)

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

W009
MEDIUM

Direct money access capability detected (payment gateways, crypto, banking).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Nov 15, 2026, 11:24 PM
Issues
2