MiniMax Multimodal (Speech + Image)

MiniMax 多模态技能 — 接入 MiniMax Token Plan 接口，语音合成（TTS/音色克隆/音色设计） 和图片生成（文生图/图生图）。使用 speech-2.8-hd（语音）和 image-01（图像）模型， 消费 Token Plan 额度。当用户提到语音合成、音色克隆、图片生成、文生图、图生...

Before installing, consider the following: - Verify the provider domains (https://api.minimaxi.com and https://api.minimax.io) are the legitimate MiniMax endpoints you expect. If unsure, contact the provider or check an authoritative homepage — the skill lists no homepage. - Treat MINIMAX_API_KEY as sensitive: the client will send it with every request and the service will charge Token Plan credits for usage (TTS, cloning, image generation). Ensure you understand billing and rate limits. - Voice cloning uploads local audio to the remote /files endpoint. Do not upload private or sensitive audio without explicit consent — this transmits user data to the provider. - The skill bundles Python scripts that import the 'requests' library but does not declare dependencies or provide an install step. Install 'requests' in your environment or run in an isolated/sandboxed environment first. - The registry metadata omits the required env vars (MINIMAX_API_KEY, MINIMAX_REGION). This is a packaging inconsistency; ask the publisher to update metadata so automated policy/permission checks can surface required credentials before installation. - The image client will download URLs returned by the API. While expected, this means the skill may fetch remote content; consider network restrictions if you run in a sensitive environment. - If you plan to use this in production or with sensitive data, request additional provenance (publisher identity, homepage, or source repo) and run the code in a controlled test environment first. If you cannot verify the provider or correct the metadata/dependency omissions, treat the skill as higher risk and avoid providing production credentials or sensitive data.