open-image

name	open-image
description	Image generation and manipulation via MCP
version	1.0.0
metadata	{"tags":["image","generation","ai","creative","mcp"],"category":"creative","related_skills":["open-creative","open-code"]}

Open Image (Image Generation MCP Integration)

Generate and manipulate images using AI-powered image generation MCP server.

When to Use

• Generate images from text descriptions
• Create visual content for presentations
• Generate mockups and prototypes
• Create marketing materials
• Generate test images for development

When NOT to Use

• Photo editing or retouching
• Vector graphics creation
• 3D modeling
• Video generation
• Real-time image processing

Prerequisites

• Image generation MCP Docker container running (docker-compose.mesh.yml includes image-mcp)
• MCP server accessible at http://image-mcp:3300
• API key configured (if required by provider)
• Agent profile with creative_generation capability

Architecture

OpenOrchestrator (plan with image generation)
  → OpenAgents (creative profile)
  → MCP call to image-mcp-server
  → Image container (generation engine)
  → Results → OpenRec (audit) → OpenBrain (observation)

Image MCP Tools

Text-to-Image Generation

image_generate(prompt: string, options?: {
  width?: number,
  height?: number,
  style?: string,
  quality?: 'standard' | 'high'
})

Generate image from text description.

Returns:

{
  "image_url": "http://image-mcp:3300/images/generated-123.png",
  "width": 1024,
  "height": 1024,
  "format": "png",
  "prompt_used": "A beautiful sunset over mountains"
}

Image Variation

image_variate(image_url: string, prompt?: string, variations?: number)

Create variations of existing image.

Returns:

{
  "variations": [
    {
      "image_url": "http://image-mcp:3300/images/var-1.png",
      "similarity": 0.85
    }
  ]
}

Image Editing

image_edit(image_url: string, edit_prompt: string, mask?: string)

Edit image based on text instructions.

Returns:

{
  "image_url": "http://image-mcp:3300/images/edited-123.png",
  "original_url": "http://image-mcp:3300/images/original.png",
  "edit_applied": "Add a red hat to the person"
}

Image to Text

image_describe(image_url: string)

Generate text description of image.

Returns:

{
  "description": "A person standing on a mountain at sunset",
  "tags": ["person", "mountain", "sunset", "nature"],
  "confidence": 0.92
}

Procedure

Basic Image Generation

1. image_generate(prompt="A professional headshot of a software developer", width=512, height=512)
2. Download generated image
3. Store in asset library
4. Log to OpenRec

Image Variation Workflow

1. image_generate(prompt="Company logo concept")
2. image_variate(image_url=result.image_url, variations=5)
3. Review variations
4. Select best option
5. Use in final design

Image Editing

1. Load existing image
2. image_edit(image_url=url, edit_prompt="Make the background blue")
3. Verify edited result
4. Save final version

Decision Rules

Scenario	Action
Need new image from scratch	Use image_generate
Need variations of existing image	Use image_variate
Need to modify existing image	Use image_edit
Need to understand image content	Use image_describe
Need high-quality output	Set quality: 'high'
Need specific dimensions	Set width and height

Pitfalls

• Don't generate images for inappropriate content
• Don't assume generated images are copyright-free
• Don't use for photorealistic human faces (ethical concerns)
• Don't generate very large images (>2048px) without testing
• Don't forget to specify dimensions if needed
• Don't use generated images without review

Verification

• Image MCP container running (docker ps | grep image-mcp)
• Can generate images from text prompts
• Can create variations
• Can edit existing images
• Generated images are valid format
• Results logged to OpenRec

Security

• Image generation requires creative_generation capability
• API keys stored securely (not in prompts)
• Generated images stored in isolated directory
• All operations logged to OpenRec
• Content filtering enabled

Cost Considerations

• Image generation may incur API costs
• Higher quality = higher cost
• Batch generation for efficiency
• Monitor usage via OpenRec

Integration with Mesh

// OpenOrchestrator plan step
{
  "goal": "Generate marketing banner image",
  "required_skills": ["open-image"],
  "approval_required": false
}

// OpenAgents dispatch
POST /v1/runs
{
  "profile": "creative",
  "goal_id": "...",
  "parameters": {
    "prompt": "Modern tech company banner with blue gradient",
    "width": 1200,
    "height": 400,
    "quality": "high"
  }
}

// OpenRec audit
{
  "type": "image.generation.completed",
  "payload": {
    "prompt": "Modern tech company banner...",
    "dimensions": "1200x400",
    "format": "png",
    "duration_ms": 3500
  }
}

Related Skills

• open-creative - General creative content generation
• open-code - Image processing and manipulation
• open-browser - Screenshot capture
• open-toolbox - Discover additional image tools