byteplus-seed-speech-tts
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents a synthesis workflow that ingests external text for processing by a remote API, creating a surface where malicious instructions could influence agent behavior.
- Ingestion points: The
textparameter in the request body examples withinSKILL.md. - Boundary markers: The documentation lacks specific instructions for using delimiters or boundary markers to isolate user-provided text from internal system logic.
- Capability inventory: The skill facilitates network operations (WebSocket and HTTP POST) to BytePlus API endpoints for audio synthesis.
- Sanitization: While the skill suggests text normalization for billing and audio quality, it does not include specific guidance for sanitizing user-provided text to prevent instruction-like content from being processed.
Audit Metadata