使用 edge-tts 生成语音并发送到飞书（语音气泡形式）。支持 XiaoxiaoNeural 等多种音色。

This skill appears to implement TTS → ffmpeg → Feishu upload correctly, but there are a few red flags to consider before installing or running it: - The script forcibly attempts to kill common browser processes (taskkill for chrome/msedge/firefox/iexplore). This behavior is not documented in SKILL.md, is unnecessary for normal operation, and can disrupt users. Inspect or remove the close_browser_processes() call if you don't want browsers terminated. - The script accepts App Secret via a CLI argument (and has a placeholder default). Passing secrets on the command line can expose them to other local users via process listings. Prefer using environment variables or a secure secret store and modify the script to read from them instead of CLI args. - Ensure the Feishu App ID/Secret you provide have the minimum required permissions (message upload/send) and are not shared broadly. The script uploads files and sends messages using your tenant token — a compromised credential could be abused. - Confirm ffmpeg installed on your system includes libopus support as required by the script. If you are comfortable reviewing/removing the browser-kill code and handling secrets securely (or running the script in a controlled environment), the skill's behavior is coherent with its description. If you cannot inspect or modify the code, treat it as risky to run on a personal or multi-user machine.

Type: OpenClaw Skill Name: feishu-voice-skill-v2 Version: 1.0.1 The skill contains an undocumented and highly irregular function `close_browser_processes` in `skill.py` that attempts to forcefully terminate all running instances of Chrome, Edge, Firefox, and Internet Explorer using the `taskkill` command. This behavior is entirely unrelated to the stated purpose of Text-to-Speech generation and Feishu integration, and it could lead to data loss or disruption of the user's environment without warning.