GPT Image 2 Prompt Architect Skill

Public OpenClaw skill and optional read-only MCP workflow for researching Image3 examples and turning rough briefs into structured GPT Image 2 prompts.

What this skill does

gpt-image-2-prompt-architect turns rough AI image ideas into cleaner GPT Image 2 prompt packs for text-to-image, reference-image editing, product photos, UI mockups, posters, social creatives, and image-to-video source frames.

It is designed for creator, ecommerce, design, and marketing workflows that need stronger subject control, readable text, product consistency, commercial composition, and faster revision loops.

Canonical links

Docs page: https://image3.org/docs/gpt-image-2-prompt-architect
Raw SKILL.md: https://image3.org/skills/gpt-image-2-prompt-architect/SKILL.md
Generator demo: https://image3.org/models/gpt-image-2
Prompt gallery: https://image3.org/prompts/gpt-image-2
Image3 Prompt MCP: https://github.com/gpt-img-2/image3-prompt-mcp
ClawHub: https://clawhub.ai/gpt-img-2/skills/gpt-image-2-prompt-architect

Best use cases

Turn a loose image idea into a structured GPT Image 2 prompt
Rewrite unstable prompts with clearer subject, layout, style, and constraints
Generate 2 to 3 focused prompt variants for rapid testing
Create ecommerce product photos, packaging mockups, lifestyle ads, detail shots, and social creatives
Prepare stable source frames, character sheets, product references, and storyboards for image-to-video workflows

Optional read-only MCP

The Image3 Prompt MCP adds live prompt search without adding generation, billing, account access, or write operations. Its tools are:

search_prompts
get_prompt
list_categories
build_prompt_brief

OpenClaw setup:

openclaw mcp add image3 \
  --command npx \
  --arg -y \
  --arg github:gpt-img-2/image3-prompt-mcp \
  --include 'search_prompts,get_prompt,list_categories,build_prompt_brief'

openclaw mcp doctor image3 --probe

Codex setup:

codex mcp add image3 -- npx -y github:gpt-img-2/image3-prompt-mcp

Submission-ready summary

Use this summary when submitting the skill to directories:

Turns rough AI image ideas into structured GPT Image 2 prompt packs, reference-image edit instructions, product photo prompts, and debugging loops for ecommerce visuals, UI mockups, posters, social creatives, and image-to-video source frames.

Submission notes

If a marketplace accepts a direct SKILL.md link, use the raw file URL above so the backlink points directly to image3.org.

If a marketplace prefers an uploaded folder, upload skills/gpt-image-2-prompt-architect.