Skip to content
Get packs

Guide · Agent skill

How to use the MiniMax AI CLI agent skill (cli-anything-minimax)

Updated · written with AI from the project's own documentation and checked against our catalogue data

The cli-anything-minimax skill teaches an AI agent to run a command-line tool that talks to MiniMax AI, a hosted AI service. Once installed, the agent can send chat prompts to MiniMax models, stream replies, and generate speech audio from text. This guide covers what the skill does, how to install it, how to trigger it with plain requests, and what it needs to run safely.

MiniMax AI CLIHKUDS/CLI-Anything · Apache-2.0

Quick start

  1. 1Check prerequisites. Confirm you have Python 3.10 or newer and a MiniMax API key from platform.minimax.io.
  2. 2Install the tool. Run the pip install command from the HKUDS/CLI-Anything repository to install cli-anything-minimax.
  3. 3Install the skill. Unzip the downloaded ZIP into ~/.claude/skills/ or a project's .claude/skills/, or upload it under Settings → Capabilities → Skills.
  4. 4Set the API key. Export MINIMAX_API_KEY or save the key with config set api_key.
  5. 5Test the connection. Run cli-anything-minimax test to confirm the API is reachable with your key.
  6. 6Ask the agent for a task. Make a natural request such as asking for a chat answer or a voiceover, and let the agent choose the command.

What is an agent skill, and what does this one do?

An agent skill is a folder containing a SKILL.md file that tells an AI agent when and how to use a tool. This skill wraps the cli-anything-minimax command-line tool, which is built on the MiniMax API.

With the skill installed, the agent can run chat completions with MiniMax-M3 (the default) or MiniMax-M2.7, stream chat responses, and synthesize speech with the tts command. The tool also manages a local config file and session history, and it can test API connectivity.

  • Chat with MiniMax LLMs, including streamed responses
  • Convert text to MP3, FLAC or PCM speech with selectable voices
  • List available voice IDs and TTS and chat models
  • Read and write local configuration, and view or clear session history
  • Check API connectivity with a test command

Install the cli-anything-minimax skill

You can add the skill by downloading the ZIP from this page and unzipping it into one of two folders. Use ~/.claude/skills/ for a personal skill available across projects, or a project's .claude/skills/ folder to limit it to that project. You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.

The skill's own install step installs the command-line tool. It requires Python 3.10 or newer and is installed from the HKUDS/CLI-Anything GitHub repository.

  • Download the ZIP from this page and unzip it into ~/.claude/skills/ or a project's .claude/skills/
  • Or upload the ZIP in the Claude apps under Settings → Capabilities → Skills
pip install git+https://github.com/HKUDS/CLI-Anything.git#subdirectory=minimax/agent-harness

Set your MiniMax API key

Every call needs a MiniMax API key, which you get from platform.minimax.io. You can set it as an environment variable for the current shell session, or save it in the tool's config file with config set. Config commands also let you check where the file lives.

  • Environment variable: export MINIMAX_API_KEY (set in your shell)
  • Config file: config set api_key, then config path to see where it is stored
export MINIMAX_API_KEY="your-api-key" # or cli-anything-minimax config set api_key "your-api-key"

Run chat and streaming commands

Use chat for a single reply and stream for output as it is generated. Add --model to switch from the default MiniMax-M3 to MiniMax-M2.7. Use --json when an agent needs parseable output.

Run --help to see all commands, or start with test to confirm your key and connection work.

cli-anything-minimax test cli-anything-minimax chat --prompt "Explain quantum computing" cli-anything-minimax chat --prompt "Quick answer" --model MiniMax-M2.7 cli-anything-minimax stream --prompt "Write a Python quicksort" cli-anything-minimax --json chat --prompt "Hello"

Generate speech audio with tts

The tts command turns text into an audio file. --text and --output are the main inputs. The default TTS model is speech-2.8-hd. Use speech-2.8-turbo for faster synthesis, and run voices to list voice IDs.

Audio controls include --speed (0.5 to 2.0), --vol (0 to 10), --pitch (-12 to 12), --sample-rate, --bitrate, --format (mp3, pcm or flac) and --channel (1 or 2). Run tts --help to see defaults.

  • --speed accepts 0.5 to 2.0
  • --vol accepts 0 to 10
  • --pitch accepts -12 to 12
  • --format accepts mp3, pcm or flac
  • --channel accepts 1 or 2
cli-anything-minimax tts --text "Hello!" --output hello.mp3 cli-anything-minimax tts --text "Fast" --model speech-2.8-turbo --voice English_Insightful_Speaker --output fast.mp3

Safety, privacy and limits

This skill runs scripts and uses the network. Prompts and text you send are transmitted to a third-party cloud service, so avoid sensitive or confidential material. Internet access is required, so the skill will not work offline. Billing applies to the provider's account.

The source's notes for agents suggest using absolute paths for TTS output files and checking return codes (0 for success, non-zero for errors), reading stderr on failure.

  • Do not send confidential text, since it goes to MiniMax's cloud service
  • Every call requires a valid MiniMax API key and an internet connection
  • Use absolute paths for --output so the agent saves files where you expect

What people use it for

  • Generate a spoken voiceover from a text file for a short video, saving an MP3 with --output
  • Ask an agent to draft or summarise content with MiniMax-M3 and return JSON it can parse
  • Compare MiniMax-M3 and MiniMax-M2.7 answers to the same prompt while evaluating models
  • Batch-style scripting of chat prompts from the terminal using the chat command
  • Produce PCM or FLAC audio by setting --format when you need a different output format

Frequently asked questions

Is the cli-anything-minimax skill free?

The source does not state pricing or licence terms for MiniMax. The skill itself is listed under Apache-2.0. Every call requires a MiniMax API key, and billing applies to the provider's account.

What do I need to install cli-anything-minimax?

You need Python 3.10 or newer, a MiniMax API key from platform.minimax.io, and internet access. The tool is installed with pip from the HKUDS/CLI-Anything GitHub repository.

How do I set my MiniMax API key?

Export MINIMAX_API_KEY in your shell, or run cli-anything-minimax config set api_key with your key. Use config path to see where the config file is saved.

Can I use this skill offline?

No. The tool needs internet access to reach the MiniMax API for both chat and text-to-speech.

Is it safe to send confidential text through this skill?

Be cautious. Prompts and text are sent to a third-party cloud service, so avoid sensitive or confidential material.

What are the alternatives to this skill?

The source does not compare this skill with other tools. If you want other LLM or speech providers, you would need to check their own skills or CLIs; this page cannot make a verified comparison.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

This skill suits developers and creators who already have a MiniMax API key and want chat or text-to-speech from the terminal or through an agent. It is a poor fit if you lack an account, need offline work, or must keep text confidential, because every request goes to MiniMax's cloud service.

Download the skill