flux-prompt-engineer

Expert prompt engineering for FLUX.1 image generation. Use when users request AI-generated images, artwork, illustrations, or visual content. Converts any visual request into optimized FLUX.1 prompts using layering, descriptive language, technical parameters, and text integration techniques.

⭐5stars🔀1forks📁View on GitHub🕐Updated Jan 11, 2026⬇️1downloads

When & Why to Use This Skill

This Claude skill provides expert-level prompt engineering specifically for FLUX.1 image generation. It transforms simple user ideas into highly detailed, layered, and technically optimized prompts to ensure superior visual outputs, photorealistic textures, and precise text rendering. By automating the transition from vague concepts to professional-grade instructions, it maximizes the creative potential of the FLUX.1 [dev] model.

Use Cases

  • Professional Photography Simulation: Generate high-fidelity, photorealistic images by applying specific camera settings (like Canon EOS R5), lighting parameters, and lens types to your visual requests.
  • Complex Scene Composition: Utilize hierarchical layering techniques to organize intricate scenes into distinct foreground, middle ground, and background elements for better spatial depth and clarity.
  • Brand and Marketing Design: Create visual assets with precise text integration by leveraging FLUX.1's advanced rendering capabilities through detailed font and typography descriptions.
  • Creative Concept Exploration: Produce three distinct prompt variations for any single idea, allowing for rapid iteration and exploration of different artistic styles, moods, and technical interpretations.
nameflux-prompt-engineer
descriptionExpert prompt engineering for FLUX.1 image generation. Use when users request AI-generated images, artwork, illustrations, or visual content. Converts any visual request into optimized FLUX.1 prompts using layering, descriptive language, technical parameters, and text integration techniques.

FLUX.1 Prompt Engineer

This skill transforms user requests into three expert-level FLUX.1 prompts, each optimized for the [dev] variant.

Core Workflow

For any image generation request:

  1. Analyze the request - Identify key visual elements, mood, style, composition
  2. Generate three variants - Create three distinct code blocks, each containing a complete prompt
  3. Apply advanced techniques - Use layering, specificity, contrast, and technical parameters from bundled references
  4. Format as code blocks - Present each prompt in a code block for easy copying

Prompt Construction Principles

Be Specific and Descriptive

Use precise language: "vibrant orange and pink sunset" not "nice sunset"

Use Hierarchical Layering

Organize complex scenes into foreground, middle ground, and background layers

Leverage Technical Parameters

Include camera settings (Canon EOS R5, 50mm lens, f/2.8, ISO 100) for photorealistic styles

Create Dynamic Scenes

Use active language: "emerging through swirling mist" not "a misty scene"

Avoid Anti-patterns

  • No Stable Diffusion syntax (e.g., ++, (emphasis))
  • No vague terms without specificity
  • No "white background" in [dev] variant (causes fuzzy outputs)
  • Keep prompts focused, not overcomplicated

Output Format

Always provide exactly three code blocks:

[Variant 1: Full detailed prompt]
[Variant 2: Alternative interpretation]
[Variant 3: Creative variation]

Advanced Techniques

For complex requests, consult these references:

  • Layered compositions: See references/layering-hierarchy.md
  • Contrasts and aesthetics: See references/contrasts.md
  • Text integration: See references/text-rendering.md
  • Technical parameters: See references/camera-settings.md
  • See-through materials: See references/materials-textures.md

Quick Examples

User: "A futuristic city"

Output:

A sprawling futuristic metropolis at twilight with towering glass skyscrapers featuring holographic advertisements. In the foreground, elevated maglev train tracks with a sleek white train passing through. The middle ground shows interconnected sky bridges between buildings with pedestrians visible as silhouettes. Background features a gradient sky transitioning from deep purple to orange, with distant communication towers topped with blinking lights. Shot on Canon EOS R5 with 24mm wide-angle lens, f/8, capturing sharp architectural details with slight motion blur on the train.
Aerial view of a neon-lit cyberpunk city at night. Foreground: rain-slicked streets reflecting pink and blue neon signs in puddles. Middle ground: dense clusters of buildings with illuminated windows creating a mosaic pattern. Background: enormous holographic advertisements projected into misty air. Atmosphere: moody and atmospheric with volumetric lighting cutting through the fog. Style: influenced by Blade Runner aesthetics with emphasis on vertical architecture and vibrant color contrasts.
A futuristic city floating above clouds, with modular buildings connected by transparent tubes. Foreground shows a landing platform with small spacecraft. Middle ground features botanical gardens suspended between structures with lush greenery visible through glass domes. Background displays the curvature of Earth with stars visible above. Lighting: soft golden hour sunlight creating warm reflections on metallic surfaces. Shot with Sony Alpha 7R IV, 50mm lens, f/4, emphasizing clean lines and utopian design philosophy.

Notes

  • FLUX.1 [dev] uses guidance scale 3.5, 50 inference steps, 1024x1024 default
  • [dev] variant excels at text rendering - include specific font descriptions when relevant
  • Avoid importing syntax from other AI tools