qwen3-tts
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or persistence mechanisms were detected in the skill instructions or examples.
- [COMMAND_EXECUTION]: The skill provides standard command-line examples for interacting with official API endpoints. Evidence: Example curl commands in SKILL.md target the authenticated Alibaba Cloud Model Studio service.
- [EXTERNAL_DOWNLOADS]: The skill refers to official software development kits and open-source models from well-known repositories. Evidence: References to the dashscope Python library and the QwenLM/Qwen3-TTS GitHub repository.
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing of user-provided content via external synthesis services, representing an ingestion surface. 1. Ingestion points: User-defined text and instructions provided in synthesis requests. 2. Boundary markers: None specified for the input text interpolation. 3. Capability inventory: Network communication with the Alibaba Cloud DashScope API. 4. Sanitization: No explicit input sanitization is prescribed in the guidance.
Audit Metadata