A professional integration skill for orchestrating MiniMax multimodal APIs, covering text, speech, video, and music generation with advanced model routing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install minimax
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install minimax using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is MiniMax?
The MiniMax skill for Openclaw Skills transforms how developers interact with the MiniMax multimodal platform. Rather than treating it as a simple API swap, this skill provides a robust framework for managing complex workflows across text, speech, and generative media. It excels at navigating the nuances between native MiniMax APIs and compatibility layers (Anthropic/OpenAI), ensuring that parameters aren't ignored and quality remains high.
By implementing this skill, agents gain the ability to handle long-running asynchronous media jobs, such as video and music generation, using proven submit-poll-fetch patterns. It establishes clear operational boundaries, providing a structured memory system in ~/minimax/ to track durable context, approved defaults, and incident reports to improve reliability over time.
MiniMax Use Cases
- Orchestrating multimodal workflows that require switching between high-speed and high-quality MiniMax-M2.5 models.
- Generating low-latency speech audio using optimized endpoints like api-uw.minimax.io.
- Managing asynchronous video and music generation queues without blocking the main execution thread.
- Implementing MCP-aware tools while maintaining strict host trust and data scope boundaries.
- Debugging complex API interactions involving malformed payloads or ignored parameters in compatibility layers.
How MiniMax Works
- Modality Identification: The skill first locks the output type (text, audio, video, or music) to determine the correct API logic and latency expectations.
- Interface Selection: It evaluates whether to use native MiniMax APIs for full feature support or compatible SDKs for existing integrations.
- Model Routing: Tasks are routed to specific model families, such as MiniMax-M2.5 or M2.1, based on the user's priority for speed versus quality.
- Execution & Polling: For media jobs, the skill initiates a submit-poll-fetch cycle, managing the asynchronous nature of generative video and music.
- Context Persistence: Successful parameters, model choices, and failure logs are saved to the local memory directory to inform future requests.
MiniMax Setup
To get started with MiniMax via Openclaw Skills, ensure you have your API key ready and follow these steps:
# Export your MiniMax API Key
export MINIMAX_API_KEY="your_api_key_here"
# Install the skill via clawhub
clawhub install minimax
# Initialize the memory structure
# This creates ~/minimax/ and necessary configuration files
Ensure curl and jq are installed on your system for optimal CLI-based troubleshooting and payload parsing.
MiniMax Data Schema & Taxonomy
The skill maintains a structured directory at ~/minimax/ to ensure durable operation of Openclaw Skills:
| File | Purpose |
|---|---|
memory.md |
Stores durable context and activation boundaries. |
routing.md |
Tracks successful model and interface choices. |
text-defaults.md |
Records model pins and SDK compatibility notes. |
speech-defaults.md |
Manages voice settings, formats, and consent notes. |
media-jobs.md |
Tracks async video/music job polling and output states. |
incidents.md |
Logs rate limits, bad prompts, and recovery strategies. |
MiniMax Advanced Features
- Multimodal Routing: Automatically selects between quality-first and speed-first model tiers based on task requirements.
- Async Job Management: Built-in logic for handling queued media generation, preventing workflow hangs during long video/music renders.
- Compatibility Shim Validation: Actively identifies when features are missing from OpenAI or Anthropic-compatible shims to avoid silent failures.
- MCP Integration: Securely orchestrates Model Context Protocol tools with explicit host review and data scope management.
- Consent-Aware Workflows: Enforces manual checkpoints before sensitive actions like voice cloning or large media uploads.
SKILL.md
Loading
Related Openclaw Skills
GitHub Private Repo SSH Routing
Fix GitHub SSH permission errors and manage multiple private repository deploy keys using intelligent host routing.
Music Playlist
An intelligent music curation and analysis skill for generating mood-based playlists and deep lyrical insights.
ELI5 / TLDR Skill
A communication-focused skill that ensures every complex AI response is followed by simplified, high-level alternatives for better accessibility.
AI Video Editor
A comprehensive AI-powered video automation skill that handles clipping, resizing, and stylistic editing through natural language prompts.
Skill Quality Checker
An automated diagnostic tool designed to evaluate the quality and performance of installed skills based on five key technical dimensions.
OpenClaw Config Guard
A deterministic configuration management tool designed to audit, backup, and safely repair OpenClaw setup files.
Featured*
