image

Generate EVOLEA brand images using reinforcement learning and iterative feedback.

Updated Dec 4, 2025
One-click install
npx skills add https://github.com/cgjen-box/evolea-website --skill image-cgjen-box
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: image
Source: https://github.com/cgjen-box/evolea-website/tree/main/.claude/skills/image-generation-rl
Command: npx skills add https://github.com/cgjen-box/evolea-website --skill image-cgjen-box

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This skill provides an RL-driven system to produce EVOLEA brand-consistent images by learning from user feedback, reducing manual trial-and-error in visual design.

Core Features & Use Cases

  • RL-based prompt refinement that improves consistency with the EVOLEA brand guide.
  • End-to-end image generation support for website hero visuals, program illustrations, and decorative assets.
  • Training, evaluation, and publishing workflow with logs and learnings to guide future iterations.

Quick Start

Provide a starting prompt such as 'Create a soft watercolor EVOLEA hero image with lavender and mint colors' to generate an initial concept.

Frequently Asked Questions about image

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How does reinforcement learning improve brand image generation?

Reinforcement learning improves brand image generation by applying iterative feedback and brand constraints to refine prompts, reducing manual trial-and-error while ensuring visual consistency.

How do I generate a brand-consistent hero image using prompts?

To generate a brand-consistent hero image, provide a descriptive starting prompt specifying style and colors, and let the RL training loop refine the output against brand constraints.

Can I use this system for decorative assets and program illustrations?

Yes, this system supports end-to-end image generation for website hero visuals, program illustrations, and decorative assets using its RL-driven training and publishing workflow.

What is the best way to maintain visual consistency across generated website visuals?

The best way to maintain visual consistency is leveraging the RL training loop, which evaluates outputs against brand constraints and logs learnings to guide future image iterations.

Does this image generation workflow require manual prompt adjustments?

Manual prompt adjustments are minimized because the system uses reinforcement learning to automatically refine prompts based on iterative feedback, reducing manual trial-and-error.

Why does the system log learnings during the image training loop?

The system logs learnings during the training loop to document evaluation results and brand constraint applications, providing actionable guidance to improve future image generation.