Skills
Navigate
skills/wlzh/skills/text-to-speech/Gen Agent Trust Hub

text-to-speech

Pass

Audited by Gen Agent Trust Hub on Oct 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes subprocess.run to invoke ffprobe for validating audio sample duration, bitrate, and codec information during voice cloning. It also supports optional post-processing by calling a secondary voice_change.py script from a related skill to apply audio filters.
  • [EXTERNAL_DOWNLOADS]: The skill performs network requests to well-known TTS providers (MiniMax and Microsoft Edge) to fetch synthesized audio. It also downloads word-level timestamp metadata from provider-supplied HTTPS URLs to support subtitle generation.
  • [INDIRECT_PROMPT_INJECTION]: The skill accepts arbitrary text input from files or standard input for conversion. It includes a ScriptParser class that sanitizes the input by identifying and removing time-stamps, background music notes, and stage directions before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 4, 2026, 04:15 AM