有声读物生成助手

Use when: 用户希望把带有 `[角色]文本` 标记的小说、剧本、故事台词转成多角色有声作品时触发。 适用于旁白、人物对白、角色 ID 已标注清楚的文本内容。Skill 会读取可编辑音色库，分析角色数量与性格特征，匹配最接近的音色，逐段调用 SenseAudio TTS，最后拼接为完整音频并以 `MEDIA...

This skill appears coherent: it needs only your SenseAudio API key to call the vendor's TTS endpoint and produces per-segment WAV output. Before installing, consider: (1) The script will try to pip-install the 'requests' package at runtime if missing — review/allow that network install on your environment. (2) The SENSEAUDIO_API_KEY you provide will be sent to the external SenseAudio API (default host api.senseaudio.cn); only supply a key you trust with this service and consider scoping/rotating it. (3) SKILL.md mentions SENSEAUDIO_API_BASE (optional) though it's not declared as required—if you need a custom endpoint, set that env var. (4) Outputs are written to the workspace outputs/ directory and local files (voice library, analysis JSON) are read/written; verify you are comfortable with that. If you want extra assurance, review scripts/main.py fully in your environment (it is included) before running, and consider running in a sandbox or with a dedicated API key with limited permissions.

Type: OpenClaw Skill Name: story-audio-adapter Version: 1.0.0 The skill bundle is a legitimate tool designed to convert structured story text into multi-role audio using the SenseAudio TTS API. The Python script `scripts/main.py` handles role parsing, heuristic voice matching, and audio concatenation using standard libraries and `ffmpeg`. While the script includes an automatic dependency installer (`pip install requests`) and executes `ffmpeg` via subprocess, these actions are transparently documented and implemented safely without shell injection risks. There is no evidence of data exfiltration, credential theft, or malicious prompt injection in `SKILL.md` or the supporting markdown files.