菜单
Crafts three structured text-to-image prompts with composition, lighting, palette, and negative constraints.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
Creator-side battle comparing the blog outline and YouTube script for the same topic.
Battle template comparing Claude and OpenAI on the same code review task using a four-axis rubric.
Compares single-lens and workflow-based finance report explanations. This is not financial advice.
Compares two models on creator planning using the YouTube content workflow.
Battle template comparing AI legal-review outputs on identical contract text. Analysis only — NOT legal advice.
Compares legal-adjacent document review outputs. This is not legal advice and must be reviewed by a qualified lawyer.
基于 SOC 职业分类
| name | ai-image-prompt-builder |
| description | Crafts three structured text-to-image prompts with composition, lighting, palette, and negative constraints. |
You are the AI Image Prompt Builder Lens. Turn the subject [[subject]] and visual style [[style]] into three production-grade prompts for text-to-image models.
Each prompt MUST include:
--no <thing>)After the prompts, list two A/B variant ideas a creator could explore next. Avoid clichés like "highly detailed, 8k".
Most prompt builders are checklist-of-words generators that produce indistinguishable outputs. This lens forces the model to commit to composition, palette, and a negative constraint — the three knobs that actually change the result across SDXL, Midjourney, and Imagen.