xAI Image Generator
Scanned@mexicanamerican
npx machina-cli add skill @mexicanamerican/xai-image-gen --openclawxai-image-gen
Generate images using xAI's Grok API (grok-imagine-image model).
Description
Production-ready CLI tool for generating images via xAI's image generation API. Supports multiple output formats, resolutions, and batch generation. Automatically outputs MEDIA: paths for OpenClaw auto-attachment.
Features:
- 🎨 Simple CLI interface:
xai-gen "<prompt>" - 🖼️ Multiple output formats: URL download, base64 encoding
- 🔢 Batch generation (multiple images per prompt)
- ⚡ Fast, pure API implementation (Pi-safe)
- 🛡️ Robust error handling with user-friendly messages
- 📎 Auto-attaches generated images in OpenClaw
- 🎯 Uses xAI's native resolution (no size parameter needed)
Installation
# Navigate to skills directory
cd ~/.openclaw/workspace/skills
# Clone or copy this skill
# (or install via clawhub when published)
# Install dependencies
pip3 install requests
# Ensure the script is executable
chmod +x xai-image-gen/xai-gen
Set your xAI API key:
export XAI_API_KEY="your-api-key-here"
Add to your shell profile (~/.bashrc, ~/.zshrc) to persist:
echo 'export XAI_API_KEY="your-api-key-here"' >> ~/.bashrc
Usage
Basic Usage
# Generate with simple prompt
xai-gen "sunset over mountains"
# Custom filename
xai-gen "cyberpunk city" --filename city.png
# Generate multiple images
xai-gen "futuristic vehicle" --n 3
# Base64 output (no download)
xai-gen "logo design" --format b64
# Verbose mode
xai-gen "space station" --verbose
Options
positional arguments:
prompt Text description of the image to generate
options:
-h, --help Show help message
--model MODEL Model name (default: grok-imagine-image)
--filename FILENAME Output filename (default: out.png)
--format {url,png,b64}
Response format: url (download), png (alias), b64 (base64)
--n N Number of images to generate (default: 1)
--verbose, -v Show detailed progress
Examples
Generate a meme:
xai-gen "dumbest trade meme: YOLO panic fail" --filename trade_meme.png
Batch generation:
xai-gen "logo variations for tech startup" --n 5
# Outputs: out_1.png, out_2.png, out_3.png, out_4.png, out_5.png
High-quality artwork:
xai-gen "photorealistic portrait of a cat astronaut" --filename cat_astronaut.png
Integration with OpenClaw
The tool outputs MEDIA: /path/to/image.png which OpenClaw automatically detects and attaches to messages. Use in agent workflows:
# In an agent skill or automation
xai-gen "chart showing Q1 sales data" --filename sales_chart.png
# → Image auto-attaches to response
API Details
- Endpoint:
https://api.x.ai/v1/images/generations - Model:
grok-imagine-image - Authentication: Bearer token via
XAI_API_KEY - Rate Limits: Subject to xAI API limits (check xAI docs)
- Timeout: 60s for generation, 30s for download
Error Handling
The tool handles common errors gracefully:
- ❌ Missing API key → Clear instructions
- ❌ Network errors → Descriptive messages
- ❌ API timeouts → Retry suggestions
- ❌ Invalid parameters → Usage hints
- ❌ File write errors → Permission checks
Requirements
- Python: 3.7+
- Dependencies:
requests - API Key: xAI API key (get from https://console.x.ai)
- Network: Internet connection required
Platform Compatibility
- ✅ Linux (tested on Raspberry Pi)
- ✅ macOS
- ✅ Windows (via WSL or native Python)
- ✅ ARM64 / ARMv7 (Pi-safe, pure API calls)
Troubleshooting
"XAI_API_KEY not found"
export XAI_API_KEY="xai-..."
"requests library not found"
pip3 install requests
Permission denied
chmod +x xai-gen
API errors
- Check API key validity
- Verify account has credits
- Check xAI status page
License
MIT License - Free to use and modify
Author
Built for OpenClaw by subagent xAI Image Gen Skill Builder
Version
1.0.0 - Initial release
Overview
This production-ready CLI creates images from text prompts using xAI's Grok API. It supports multiple output formats, batch generation, and automatically outputs MEDIA: paths for OpenClaw integration, making image attachment seamless.
How This Skill Works
The tool calls xAI's Grok-imagine-image model via the /v1/images/generations endpoint using a Bearer token from XAI_API_KEY. It can return URLs, PNGs, or base64 data and can generate multiple images per prompt in a single run. It also outputs MEDIA: paths so OpenClaw can auto-attach the generated images.
When to Use It
- You need a quick image generated from a descriptive prompt for social posts or demos.
- You want multiple variants of a concept in a single run (batch generation).
- You need different output formats (URL, PNG, or base64) to fit downstream tools.
- You rely on OpenClaw and require automatic MEDIA attachments from generated images.
- You’re building AI art assets or demos with a simple, scriptable CLI workflow.
Quick Start
- Step 1: Set your API key in your environment: export XAI_API_KEY="your-api-key"
- Step 2: Run a prompt to generate images: xai-gen "sunset over mountains"
- Step 3: (Optional) specify --n for batch or --format to choose url/png/b64; rely on MEDIA: paths for OpenClaw attachment
Best Practices
- Craft concise, specific prompts to guide the Grok model toward the desired style and content.
- Use --n to generate multiple variants and select the best result.
- Ensure the XAI_API_KEY environment variable is set and valid before running.
- Choose the output format that best suits your integration (url for downloads, b64 for embedding).
- Rely on the MEDIA: outputs to verify OpenClaw auto-attachment works as expected.
Example Use Cases
- Generate a meme with a custom filename: xai-gen "dumbest trade meme: YOLO panic fail" --filename trade_meme.png
- Batch logo variations for a startup: xai-gen "logo variations for tech startup" --n 5
- Produce a photorealistic cat astronaut portrait: xai-gen "photorealistic portrait of a cat astronaut" --filename cat_astronaut.png
- Output base64 for embedding in chat: xai-gen "logo design" --format b64
- Create a data chart image with OpenClaw attachment: xai-gen "chart showing Q1 sales data" --filename sales_chart.png
