A powerful multimodal skill for generating and editing AI images and videos via the MuleRouter and MuleRun APIs.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install mulerouter
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 mulerouter 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 MuleRouter AI Multimodal Generation?
This skill provides a comprehensive interface for interacting with MuleRouter and MuleRun multimodal APIs. It allows developers to integrate advanced Text-to-Image, Image-to-Video, and video editing capabilities directly into their AI agent workflows. By leveraging state-of-the-art models such as Wan2.6, Midjourney, and Sora2, this Openclaw Skills contribution simplifies complex generative tasks through a unified CLI interface.
The integration handles the heavy lifting of API communication, model discovery, and parameter validation. It features robust support for local file handling, automatically converting images to base64 for secure and efficient processing. Whether you are building automated content creation pipelines or interactive creative tools, this skill provides the necessary building blocks for high-fidelity media generation.
MuleRouter AI Multimodal Generation Use Cases
- Automating video content creation using Text-to-Video models.
- Transforming existing images into dynamic videos using Image-to-Video workflows.
- Generating high-quality marketing assets with SOTA Image-to-Image models.
- Rapidly prototyping visual concepts within an AI agent environment.
- Building custom video editing tools that leverage keyframe interpolation and VACE.
How MuleRouter AI Multimodal Generation Works
- The skill validates environment variables, specifically the MULEROUTER_API_KEY and endpoint configurations.
- Dependencies are managed and installed via uv to ensure a clean, isolated environment.
- Users can query the API to discover available models and their specific technical tags like [SOTA].
- The skill inspects model-specific parameters to provide accurate CLI options for generation.
- Upon execution, local image files are validated and processed before being sent to the remote API.
- The generated media URLs are returned for further processing or storage.
MuleRouter AI Multimodal Generation Setup
Ensure Python 3.10+ and uv are installed. Set up your environment variables in a .env file:
MULEROUTER_API_KEY=your_api_key_here
MULEROUTER_BASE_URL=https://api.mulerouter.ai
Install dependencies and verify the setup:
uv sync
uv run python -c "from core.config import load_config; load_config(); print('Configuration OK')"
MuleRouter AI Multimodal Generation Data Schema & Taxonomy
The skill organizes its operations through a structured directory layout and metadata-driven model discovery.
| Component | Description |
|---|---|
| scripts/ | Contains general utility scripts like list_models.py. |
| models/ | Organized by provider and model type (e.g., alibaba/wan2.6-t2v). |
| core/ | Logic for configuration loading and API communication. |
| .env | Local storage for sensitive API credentials and base URLs. |
Output Format: Results typically return a JSON object containing the job_id, status, and a list of output URLs for the generated media.
MuleRouter AI Multimodal Generation Advanced Features
- Model filtering using technical tags (e.g., --tag SOTA) to programmatically select high-performance models within Openclaw Skills.
- Integrated parameter discovery (--list-params) to dynamically adapt to model updates without changing code.
- Automatic base64 conversion for local image paths to bypass CLI character limits.
- Multi-site support for switching between MuleRouter and MuleRun infrastructures.
- Support for complex video operations including VACE and keyframe interpolation.
SKILL.md
Loading
Related Openclaw Skills
mulerouter
A comprehensive multimodal interface for generating and editing AI-driven images and videos through the MuleRouter and MuleRun APIs.
Team Manager Skill
A professional automation skill designed to manage team daily routines, morning briefings, and end-of-day reporting cycles.
FunPay Assistant Skill
A specialized automation skill for FunPay sellers to manage customer inquiries and monitor account activity in real-time.
Seithar Intel
A personal cyber threat intelligence analyst that monitors global security feeds and delivers interest-calibrated briefings and deep-dives.
Maven Central Publish
A comprehensive toolkit and workflow for publishing Java and Kotlin libraries to Maven Central via the modern Sonatype Central Portal.
Home Assistant Skill
A powerful Python-based bridge for controlling Home Assistant devices, sensors, and scenes via a command-line interface.
METADATA
Featured*
