MiniMax TTS 国内版

调用MiniMax语音合成API，支持中文多音色、高质量文本转语音，提供流式和非流式音频输出。

Security Scan

The skill appears to implement a TTS client that matches its description, but there are inconsistencies (registry metadata vs SKILL.md) and a suspicious API-host/URL mismatch that warrant cautious review before use.

Purpose & Capability

The skill's name/description (MiniMax TTS) align with the code and SKILL.md: it calls a remote TTS API, accepts text and voice/model parameters, lists voices, and writes audio output. However, registry metadata claims no required env vars or primary credential while SKILL.md and the script clearly require MINIMAX_API_KEY—this metadata inconsistency is unexpected and reduces trust.

Instruction Scope

The runtime instructions and the included script stay within the stated purpose: they call TTS endpoints, list voices, decode audio and write an output file. The script does not read arbitrary user files, other environment variables, or post data to unknown endpoints beyond the declared API_BASE. It does print and write the generated audio file to disk (expected behavior for a CLI TTS tool).

Install Mechanism

There is no install spec (instruction-only plus an included Python script). That is low-risk compared to remote downloads. The SKILL.md asks for python3 and the requests package; the script imports requests. No archives or external installers are used.

Credentials

The script and SKILL.md require a single API credential (MINIMAX_API_KEY), which is proportionate for a third-party TTS API. However, the registry metadata claims no required env variables or primary credential—this mismatch is suspicious. No other secret env vars are requested or accessed by the script.

Persistence & Privilege

The skill does not request persistent or elevated privileges: always is false, it does not modify other skills or system-wide settings, and it only writes output audio files in the current working directory (or specified path).

What to consider before installing

This skill mostly behaves like a normal TTS client, but review the following before installing or using it: - Verify the API endpoint: the SKILL.md homepage references platform.minimax.io, but the script calls https://api.minimaxi.com (note the extra 'i' in minimaxi). That could be a harmless typo or point to an unintended/malicious host; confirm the correct API domain with the vendor/documentation before providing an API key. - Metadata mismatch: the registry entry claims no required env vars, while SKILL.md and the script require MINIMAX_API_KEY. Prefer skills whose published metadata matches their runtime requirements. - Limit exposure: only supply an API key that you control and can revoke; do not reuse broad-scoped or production credentials. Consider creating a limited/test account on the provider for initial use. - Code review: the script decodes hex audio and writes a local file (expected). It imports subprocess but does not use it—likely harmless but worth noting. If you have doubts, run the script in an isolated environment (container/VM) and inspect network traffic to confirm it talks only to the intended API host. - If you cannot confirm the correct API host or the publisher identity, do not provide credentials or use this skill in sensitive environments.

MiniMax TTS Skill

调用 MiniMax TTS API 生成语音。

配置

设置环境变量：

export MINIMAX_API_KEY="your-api-key"

使用方式

命令行

python3 ~/.openclaw/workspace/skills/minimax-tts/scripts/tts.py "要转语音的文本"

参数选项

参数	说明	默认值
--text	要转语音的文本	必填
--model	模型	speech-2.8-turbo
--voice	音色ID	Chinese_Male_Adult
--speed	语速	1.0
--format	音频格式	mp3
--output	输出文件	output.mp3

示例

# 基本用法
python3 ~/.openclaw/workspace/skills/minimax-tts/scripts/tts.py "你好世界"

# 指定音色和模型
python3 ~/.openclaw/workspace/skills/minimax-tts/scripts/tts.py "你好世界" --voice Chinese_Female_Adult --model speech-2.8-hd

# 保存到指定文件
python3 ~/.openclaw/workspace/skills/minimax-tts/scripts/tts.py "测试语音" --output test.mp3

可用音色

调用 get_voice API 获取当前账号下所有音色：

python3 ~/.openclaw/workspace/skills/minimax-tts/scripts/tts.py --list-voices

常见系统音色：

• Chinese_Male_Adult - 中文男声
• Chinese_Female_Adult - 中文女声
• English_Male_Adult - 英文男声
• English_Female_Adult - 英文女声

支持的模型

模型	特点
speech-2.8-hd	最高质量，40+语言
speech-2.8-turbo	低延迟
speech-2.6-hd	高相似度
speech-2.6-turbo	性价比高