An advanced OpenClaw skill for optimizing AI-generated images through prompt alignment, logical analysis, and iterative refinement.
一个基于提示词对齐、逻辑分析与多轮迭代优化的 AI 图像优化 Skill。(openclaw)
- Prompt + image joint reasoning / Prompt + Image 联合分析
- Detect issues from prompt or generation / 自动识别问题来源(Prompt / 生成)
- Iterative optimization loop (max 3 rounds) / 多轮优化(最多 3 轮)
- Score-based stopping system / 评分系统控制停止
- Automatic prompt refinement / 自动优化 Prompt
- Multi-model support (OpenAI / Gemini / Doubao / SD / Imagen / FLUX / Wan) / 多模型支持(OpenAI / Gemini / 豆包 / SD / Imagen / FLUX / 万相)
Input: prompt
Output: optimized prompt + improved image result
输入:prompt
输出:优化后的 prompt + 更合理的图像结果
Input: image
Output: issue analysis + inferred original prompt + repair suggestions
输入:image
输出:问题分析 + 推测原始 prompt + 修复建议
Input: prompt + image
Output: alignment analysis + prompt optimization + multi-round optimization result
输入:prompt + image
输出:对齐分析 + Prompt 优化 + 多轮优化结果
Analyze -> Align -> Optimize Prompt -> Generate -> Evaluate -> Loop
分析 -> 对齐 -> 优化 Prompt -> 生成 -> 评估 -> 循环
OpenClaw / GPT / Gemini / Doubao / Stable Diffusion / DALL-E / Imagen / FLUX
~/.openclaw/workspace/skills/
Input: a penguin flying in the sky
Output: detect semantic issue -> optimize prompt -> generate a more realistic version
输入:a penguin flying in the sky
输出:检测语义问题 -> 优化 Prompt -> 生成更合理版本
MIT License
Copyright (c) 2026 Clement-2333
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files...