← Back to Skills Marketplace
133
Downloads
0
Stars
0
Active Installs
7
Versions
Install in OpenClaw
/install minimax-tts-plus
Description
MiniMax TTS skill (enhanced). Multi-agent voice support (each agent can select a unique voice written in SOUL.md), native voice message for Telegram (MP3) an...
Usage Guidance
This skill appears to do what it claims: call MiniMax's API to produce audio and optionally send it to Telegram or prepare Feishu-compatible files. Before installing, consider the following:
- MINIMAX_API_KEY is required and will be used to authenticate to https://api.minimaxi.com — only provide a key you trust to be used for TTS. Do not reuse high-privilege keys elsewhere.
- Telegram credentials (TELEGRAM_BOT_TOKEN, TELEGRAM_TARGET) are only needed if you want the skill to send messages. If you do not want automatic sending, leave them unset; the skill supports a --generate-only mode.
- The registry metadata marks Telegram vars as required although the SKILL.md and scripts treat them as optional — treat the SKILL.md behavior as authoritative and verify before supplying Telegram tokens.
- The scripts load a .env file and export its entries into the environment. Keep .env private and out of version control (setup.txt already warns about this).
- The python script and shell wrapper use workspace paths (~/.openclaw/workspace). Check those directories for generated files and remove audio files if you do not want them retained.
If you need higher assurance, you can inspect/verify the two included files (scripts/tts.py and scripts/tts-xiaoye.sh) locally before supplying secrets. If you do not intend to send to Telegram/Feishu, omit the related env vars to reduce blast radius.
Capability Analysis
Type: OpenClaw Skill
Name: minimax-tts-plus
Version: 1.2.3
The skill provides legitimate Text-to-Speech functionality using the MiniMax API and supports delivery to Telegram and Feishu. However, it is classified as suspicious due to a potential file exfiltration vulnerability in `scripts/tts-xiaoye.sh`. The script passes the `CAPTION` and `TARGET` variables directly to `curl -F` without sanitization; if these variables are manipulated (e.g., via prompt injection) to start with the `@` character, `curl` will attempt to read and upload local files such as the `.env` file (containing API keys) or SSH keys. While the behavior aligns with the stated purpose, this lack of input validation in a shell script that performs network requests poses a significant security risk.
Capability Assessment
Purpose & Capability
The skill's name and description (MiniMax TTS with Telegram/Feishu delivery) match the included scripts: python code calls the MiniMax TTS API and the shell wrapper handles generation, optional ffmpeg transcode and Telegram delivery. One mismatch: the registry-level "Required env vars" lists TELEGRAM_BOT_TOKEN and TELEGRAM_TARGET as required, whereas SKILL.md and the scripts treat those as optional (only needed for sending). This is likely a metadata error rather than functional misalignment.
Instruction Scope
Instructions and scripts only read credentials from .env or environment variables, write audio files under the OpenClaw workspace (~/.openclaw/workspace/generated/*), call the MiniMax API (api.minimaxi.com), optionally call Telegram's Bot API, and use ffmpeg for conversion. The scripts do not attempt to read unrelated system files or send arbitrary data to unknown endpoints.
Install Mechanism
No install spec is provided (instruction-only plus included scripts). Required binaries are standard (python3, ffmpeg). No downloads from third-party URLs or archive extraction occur during install.
Credentials
The primary credential (MINIMAX_API_KEY) is appropriate. Telegram credentials are relevant only if you want the skill to send messages; SKILL.md marks them optional, but registry metadata lists them as required — this inconsistency should be corrected. The scripts load and export .env entries, so any secrets placed in the .env will be exported into the script environment (expected behavior but worth noting).
Persistence & Privilege
The skill does not request permanent system-wide presence (always:false) and does not modify other skills or agent configuration. It only writes generated audio files under the OpenClaw workspace and reads a local .env for configuration.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install minimax-tts-plus - After installation, invoke the skill by name or use
/minimax-tts-plus - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.2.3
Fix metadata: add TELEGRAM_BOT_TOKEN/TELEGRAM_TARGET to env list (optional). Add API endpoint note (api.minimaxi.com). Clarify Telegram vars are optional. Update setup.txt with clearer instructions.
v1.2.2
Rename env.example to setup.txt (clawhub filters env* files from publish)
v1.2.1
Fix: env.example now published (was hidden due to dotfile skip). SKILL.md updated with correct path.
v1.2.0
Security fix: remove hardcoded bot token, credentials now read from .env file only. No tokens stored in shell scripts.
v1.1.1
Add --list-voices to tts-xiaoye.sh; document how to list voices in SKILL.md
v1.1.0
Internationalization: bilingual SKILL.md (EN/CN), English comments in all scripts
v1.0.0
多 Agent 音色、多渠道原生语音条(飞书 OGG + Telegram MP3)
Metadata
Frequently Asked Questions
What is Minimax Tts Cn?
MiniMax TTS skill (enhanced). Multi-agent voice support (each agent can select a unique voice written in SOUL.md), native voice message for Telegram (MP3) an... It is an AI Agent Skill for Claude Code / OpenClaw, with 133 downloads so far.
How do I install Minimax Tts Cn?
Run "/install minimax-tts-plus" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Minimax Tts Cn free?
Yes, Minimax Tts Cn is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Minimax Tts Cn support?
Minimax Tts Cn is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Minimax Tts Cn?
It is built and maintained by vshen009 (@vshen009); the current version is v1.2.3.
More Skills