voice-ai-development

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt includes code examples that assign API keys inline (e.g., OPENAI_API_KEY = "sk-...") and shows embedding them directly into headers/clients, which encourages requesting and pasting secret values verbatim into generated code or commands.

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

• Third-party content exposure detected (high risk: 0.90). The skill ingests untrusted, user-generated content via Vapi webhooks (e.g., event["transcript"] and function-call events) and Deepgram live transcription (transcribe_stream calls handle_user_input), and those transcripts/events are explicitly used to drive behavior and tool/function calls, so third-party content can materially influence agent actions.