Agent skill

generate-image

Generate images using xAI's Grok Imagine through the fal.ai MCP gateway. This skill enables text-to-image generation with detailed prompts for photorealistic, artistic, and stylized outputs.

View SKILL.md on GitHub Repository
Stars 1
Forks 0

Install this agent skill to your Project

npx add-skill https://github.com/richardanaya/agent-skills/tree/main/.agents/skills/generate-image

SKILL.md

Generate Image Skill

Description

Generate images using xAI's Grok Imagine through the fal.ai MCP gateway. This skill enables text-to-image generation with detailed prompts for photorealistic, artistic, and stylized outputs.

When to Use

Use this skill when the user wants to:

  • Generate images from text descriptions
  • Create illustrations, artwork, or visual content
  • Use AI image generation tools
  • Produce concept art, portraits, scenes, or abstract imagery

Prerequisites

Ensure MCP servers are connected and the fal.ai server is available:

javascript
mcp({})  // Check server status

Input Schema

Parameter Type Required Default Description
prompt string Yes - Text description of the desired image
num_images integer No 1 Number of images to generate
aspect_ratio AspectRatioEnum No "1:1" Aspect ratio of the generated image
output_format OutputFormatEnum No "jpeg" Format of the generated image
sync_mode boolean No false If true, returns media as data URI

AspectRatioEnum Values

Value Description
"2:1" Ultra-wide panorama
"20:9" Cinematic ultra-wide
"19.5:9" Mobile ultra-wide
"16:9" HD widescreen (video standard)
"4:3" Classic display ratio
"3:2" Photo print ratio
"1:1" Square (default)
"2:3" Portrait photo
"3:4" Portrait display
"9:16" Mobile vertical
"9:19.5" Mobile tall
"9:20" Mobile ultra-tall
"1:2" Vertical panorama

OutputFormatEnum Values

Value Description
"jpeg" JPEG format (default, best compression)
"png" PNG format (lossless, transparency support)
"webp" WebP format (modern web optimized)

Output Schema

Field Type Description
images list<ImageFile> Array of generated image objects
revised_prompt string The enhanced prompt used by the model

ImageFile Object

Field Type Description
url string The URL of the generated image

Tools

1. Search for Image Generation Models

First, verify the available image generation models:

javascript
mcp({
  search: "image generation",
  server: "fal"
})

Or specifically search for Grok:

javascript
mcp({
  search: "grok",
  server: "fal"
})

2. Generate Image

Use the fal_run_model tool with the xai/grok-imagine-image endpoint:

javascript
mcp({
  tool: "fal_run_model",
  server: "fal",
  args: {
    "endpoint_id": "xai/grok-imagine-image",
    "input": {
      "prompt": "Your detailed image description here",
      "num_images": 1,
      "aspect_ratio": "1:1",
      "output_format": "jpeg"
    }
  }
})

3. Display the Image

After generation, display the image to the user:

javascript
show_image({
  url: "https://v3b.fal.media/files/.../image.jpg"
})

Or for local files:

javascript
show_image({
  path: "/path/to/local/image.png"
})

Example Request

json
{
  "prompt": "Abstract human silhouette, golden particles ready to burst outward representing joy, data visualization style, emotional expression through particles, artistic scientific",
  "num_images": 1,
  "aspect_ratio": "1:1",
  "output_format": "jpeg"
}

Example Response

json
{
  "images": [
    {
      "url": "https://v3b.fal.media/files/b/0a8b90b7/9avg_nKJmcVinjQHJR_Ja.jpg"
    }
  ],
  "revised_prompt": ""
}

Best Practices

Writing Effective Prompts

  1. Be specific: Include subject, style, lighting, mood, and quality descriptors
  2. Use artistic references: "cinematic lighting", "photorealistic", "oil painting", "anime style"
  3. Specify technical quality: "8k resolution", "highly detailed", "sharp focus"
  4. Describe composition: "close-up portrait", "wide landscape", "aerial view"
  5. Include mood/atmosphere: "dramatic", "serene", "futuristic", "vintage"

Example Prompts by Category

Portraits:

Portrait of a wise elderly wizard with flowing white beard, wearing intricate 
blue robes embroidered with silver stars, mystical library background, warm 
candlelight, highly detailed face, 8k, photorealistic

Sci-Fi:

Futuristic cyberpunk city at night, neon signs in Japanese, flying cars, rain 
streets reflecting lights, towering holographic advertisements, blade runner 
style, cinematic composition, 8k quality

Nature:

Majestic snow-capped mountains at golden hour, crystal clear alpine lake with 
reflection, pine forest in foreground, dramatic clouds, landscape photography 
style, ultra detailed, 8k resolution

Abstract:

Abstract human silhouette, golden particles ready to burst outward representing 
joy, data visualization style, emotional expression through particles, 
artistic scientific

Characters (e.g., Ferengi):

A Star Trek Ferengi alien, humanoid with large ears, wrinkled forehead, sharp 
teeth grin, wearing traditional Ferengi business attire with dark tunic and 
collared shirt, standing in a sci-fi alien marketplace, detailed facial features, 
dramatic lighting, cinematic quality, 8k resolution, photorealistic

Complete Workflow Example

Here's the full workflow for generating and displaying an image:

javascript
// Step 1: Generate the image
const result = await mcp({
  tool: "fal_run_model",
  server: "fal",
  args: {
    "endpoint_id": "xai/grok-imagine-image",
    "input": {
      "prompt": "A serene Japanese zen garden with raked sand patterns, moss-covered stones, cherry blossom tree in bloom, morning mist, tranquil atmosphere, traditional wooden pagoda in background, ultra detailed, 8k, photorealistic",
      "num_images": 1,
      "aspect_ratio": "16:9",
      "output_format": "jpeg"
    }
  }
});

// Step 2: Extract the image URL
const imageUrl = result.result.images[0].url;

// Step 3: Display to user
show_image({
  url: imageUrl
});

Troubleshooting

Issue Solution
Model not found Run mcp({ search: "grok", server: "fal" }) to verify endpoint
Connection error Check MCP status with mcp({}) and reconnect
Poor image quality Add more specific descriptors to prompt
Wrong aspect ratio Use "1:1", "16:9", "9:16", "4:3", or "3:4"
Generation timeout Try simpler prompt or check fal.ai status
Sync mode not working Set "sync_mode": true for data URI output

Alternative Models

If Grok Imagine is unavailable, search for alternatives:

javascript
mcp({
  search: "flux",
  server: "fal"
})

Popular alternatives:

  • fal-ai/flux-2-pro - High quality, detailed images
  • fal-ai/flux/dev - Fast generation, good quality
  • fal-ai/flux-1/schnell - Fastest inference

Notes

  • Generated images are hosted temporarily on fal.media
  • Image URLs are valid for a limited time (typically 24-48 hours)
  • Save images locally if permanent storage is needed
  • Be mindful of content policies when generating images
  • The revised_prompt field may contain an enhanced version of your prompt

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

richardanaya/agent-skills

bd-what-to-work-on-next

Determine what to work on next using the bd (beads) issue tracker by analyzing open issues, dependencies, and ready work

1 0
Explore
richardanaya/agent-skills

bd-create-new-product

Create a new product with a structured chain of related issues using the bd (beads) issue tracker - sets up epics, tasks, and dependencies for product development

1 0
Explore
richardanaya/agent-skills

pubmed-metaanalysis

Search PubMed for meta-analyses on a given medical topic using NCBI E-utilities API

1 0
Explore
richardanaya/agent-skills

interact-with-browser

a CLI for intearcting with browser

1 0
Explore
richardanaya/agent-skills

beautiful-reader

Use when the user wants to read some markdown in a beautiful way

1 0
Explore
richardanaya/agent-skills

spec-driven-development

Complete workflow for LLM coding agents: initial setup of the template system, adding new features, creating modules/components, using the Constitution as the single integrator, and documenting major architecture changes with ADRs.

1 0
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results