text-to-speech
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill communicates with api.heygen.com and resources on heygen.ai, which are well-known services for AI video and audio generation. All network operations are directed to official service domains.
- [CREDENTIALS_UNSAFE]: The skill correctly instructs users to manage sensitive API keys through environment variables (HEYGEN_API_KEY) rather than hardcoding them in scripts, following industry best practices for secret management.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted text data for conversion to audio. 1. Ingestion points: The 'text' parameter in the text_to_speech function found in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: HTTP requests to HeyGen API endpoints using requests or fetch. No file system access or shell command execution. 4. Sanitization: None described for the input text. The risk is considered negligible as the output is a hosted media file and the skill lacks exploitable local capabilities.
Audit Metadata