Free voice from Comfy UI + Qwen3 TTS
PassAudited by ClawScan on May 1, 2026.
Overview
This appears to be a benign local text-to-speech helper, but it will start and control a local ComfyUI instance and save/send generated audio.
Before installing, make sure your local ComfyUI shortcut, Qwen3 model, and custom TTS node are trusted. The skill is intended to start ComfyUI, submit the text you provide to the local service, save an MP3, and send that generated audio as a voice message.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The skill may open a local program on the user's machine before generating audio.
The skill instructs the agent to run a local command to start ComfyUI. This is purpose-aligned for a local ComfyUI TTS workflow, but users should verify the shortcut points to the expected trusted application.
cmd /c start "" "C:\Users\user\Desktop\ComfyUI.lnk"
Confirm the ComfyUI shortcut and local installation are trusted before using the skill.
Using the skill can make the local ComfyUI service generate audio and write an MP3 file.
The skill submits a workflow to a local ComfyUI HTTP API and then checks its history. This is expected for the stated purpose, but it means the skill can trigger local model work and file output through that service.
Отправить POST-запрос на `http://localhost:8000/prompt`
Use it only with a trusted local ComfyUI instance and review the text to be converted before sending.
The safety of generation depends partly on the user's local ComfyUI, plugin, and model installation.
The skill relies on external local components, including a custom ComfyUI node and Qwen3 model, but does not install or verify them. This is normal for an instruction-only integration, yet their provenance matters.
ComfyUI с установленным плагином `AILab_Qwen3TTSVoiceDesign_Advanced`
Install ComfyUI nodes and models only from trusted sources and keep them updated.