deapi

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

• Third-party content exposure detected (medium risk: 0.30). The workflow consumes media links (YouTube, X, Twitch, Kick, TikTok) via --url or source inputs in transcribe and describe commands, where outsider-submitted content (such as a public YouTube video or TikTok link) is read and transcribed/described.

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

• Direct money access detected (high risk: 0.70). The skill instructs the agent to operate a helper script (scripts/deapi.py) and MCP tools that execute paid media generation jobs and submit actual financial charges against user accounts (using DEAPI_API_KEY and pricing/cost limits). This constitutes direct financial operations (purchases/charges). These routes rely on API keys and service-side limits/authentication rather than independent execution-layer access controls validating each financial transaction independently of the agent's provided credentials. Therefore, the direct operations lack established independent access control.