dialogue-audio

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill defines several commands using the infsh CLI to generate audio and merge media files. Examples include infsh app run falai/dia-tts for dialogue synthesis. All command executions are restricted to the infsh tool via the allowed-tools policy.
  • [EXTERNAL_DOWNLOADS]: The documentation points to the inference-sh/skills@agent-tools package for CLI setup. This is a standard dependency for accessing the platform's features.
  • [PROMPT_INJECTION]: The skill processes user-supplied text for audio synthesis, which creates an indirect prompt injection surface. 1. Ingestion points: The prompt field in the shell command input. 2. Boundary markers: None provided. 3. Capability inventory: Execution of infsh commands in Bash. 4. Sanitization: None present. This surface is inherent to text-to-speech functionality and is documented here for completeness.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 12:10 PM