etoro

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt asks the agent to request the user's Public API Key and User Key and provides curl examples showing those keys placed directly into request headers, which would require embedding secret values verbatim into generated commands or responses.

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

• Third-party content exposure detected (high risk: 0.80). The skill documents and exposes public social endpoints (e.g., GET /feeds/instrument/{marketId}, GET /feeds/user/{userId}, and public watchlists like GET /watchlists/public/{userId}) that fetch user-generated eToro feed/watchlist content which the agent can read and act on, creating a risk of indirect prompt injection from untrusted third-party posts.

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

• Direct money access detected (high risk: 1.00). The skill explicitly interfaces with the eToro API and includes endpoints and request bodies for trade execution (market open orders by amount/units, market close orders, limit orders), canceling orders, and real vs demo trading. It requires API keys with Write permissions and environment selection (Real vs Virtual), so it is specifically designed to execute financial transactions (place/cancel/close trades).