TTS
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text strings to generate audio output. This represents an ingestion surface where untrusted data enters the agent context.
- Ingestion points: The
inputparameter inzai.audio.tts.create()and CLI--inputflags (SKILL.md, tts.ts) accept arbitrary text. - Boundary markers: No explicit delimiters or warnings to ignore instructions embedded in the processed text are implemented.
- Capability inventory: The skill utilizes
fs.writeFileSync(SKILL.md, tts.ts) for local file system access to save audio data. - Sanitization: Basic text preparation is suggested for pronunciation, but no sanitization against prompt injection is provided.
Audit Metadata