Xiaomi-MiMo-V2-TTS

小米 MiMo V2 TTS 文字转语音模型（官网目前免费）。支持中文/英文，内置情感风格（开心/悲伤/生气）、角色扮演（孙悟空/林黛玉）、方言（东北话/四川话/粤语/河南话）、语速控制及唱歌能力。mp3/opus 格式可直接发送至微信/飞书。 **配置（必需）**：在 `openclaw.json` 的 `sk...

What to consider before installing: - The code and docs implement TTS and will send your text and an API key to https://api.xiaomimimo.com. Only provide MIMO_API_KEY if you trust that endpoint and the skill author. - The registry metadata did NOT declare the required MIMO_API_KEY or the openclaw.json config access — this omission is suspicious and means your agent UI may not warn you about the credential the skill needs. - The script will attempt to read ~/.openclaw/openclaw.json to load the key (and checks a legacy path). Inspect that file first — it can contain other tokens/config — and avoid storing unrelated secrets there if you enable this skill. - Examples reference /root/.openclaw paths; ensure the expected installation path matches your environment and that hard-coded paths won’t expose files unintentionally. - The skill requires network access and may perform multiple retries; review the endpoint hostname and consider running the script locally in a sandbox or with a throwaway API key before enabling in production. - Installing ffmpeg via apt Homebrew requires elevated privileges on some systems; be careful granting those. If you want to proceed: (1) review the full scripts/tts.py source yourself (it’s included), (2) add only the minimal API key needed, (3) keep openclaw.json free of other secrets or avoid relying on that file by setting MIMO_API_KEY as an environment variable, and (4) run first in an isolated environment. If you need, I can highlight the exact lines in the script that read openclaw.json and send the API request.

Type: OpenClaw Skill Name: xiaomi-mimo-v2-tts Version: 2.0.0 The skill provides legitimate text-to-speech functionality using the Xiaomi MiMo V2 API. The core logic in scripts/tts.py handles API communication, audio format conversion via ffmpeg, and configuration management by reading the skill's own API key from the standard OpenClaw configuration file (~/.openclaw/openclaw.json). The behavior is well-documented in SKILL.md and README.md, and all network, file, and shell operations are strictly aligned with the stated purpose of audio generation and processing.