sellix

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs the agent to call Sellix via Membrane (e.g., "membrane action list", "membrane request CONNECTION_ID /path/to/endpoint"), which fetches and ingests untrusted, user-generated Sellix data/APIs that the agent is expected to read and that can influence subsequent actions.

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

• Direct money access detected (high risk: 1.00). The skill integrates directly with Sellix — an e-commerce/payment platform — and explicitly exposes Payments and Payouts objects, Orders, and API proxying. It documents running actions and sending arbitrary proxied HTTP requests (POST/DELETE/etc.) through Membrane with authentication handled for you. Those capabilities allow creating/managing payments and issuing payouts (i.e., moving funds), so this is a specific financial execution integration, not a generic tool.