Skip to content
Get packs

GPT Image 2 Generator

Guides an agent to write image prompts and, when enabled, generate or edit images via GPT Image 2.

Source Download skill
Repository
ConardLi/garden-skills
Stars
13,484
Licence
MIT
Files
15 files

What it does

This skill teaches an AI agent to create images with GPT Image 2 using structured prompt templates for posters, UI mockups, product visuals, infographics, academic figures and more. It first checks which mode it is running in. If an OpenAI-compatible API key and the Garden switch are set, it can call bundled scripts to generate or edit images and save them to the workspace. Otherwise it falls back to using the host agent's own image tool, or simply writes a high-quality prompt for the user to paste into another image tool.

Great for

  • Designers who want consistent, detailed prompts for posters, UI mockups or product shots
  • Marketers producing social, landing page or e-commerce visuals with structured templates
  • Researchers or students who need clean academic diagrams and figures drafted from prompts

Skip it if

  • Users who need guaranteed image output without an API key or image tool, since in prompt-only mode the agent cannot render images
  • Generating realistic images of real people or celebrities, which the templates allow (e.g., a livestream host can be a named celebrity), so review for likeness and consent issues
  • Sensitive or regulated visuals that need verified accuracy, because prompt templates do not check factual correctness

Try asking

  • “Write a prompt for a live-commerce product UI with a host, chat area and product card, then generate it”
  • “Edit my photo at assets/source.png to replace the background with a studio scene”
  • “Create a retro skeuomorphic icon set and save the prompt for reuse”
  • “Draft a graphical abstract for my research method pipeline”
Step-by-step guide: how to use GPT Image 2 GeneratorInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 15 original files, the MIT licence and an attribution note.

    gpt-image-2.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip gpt-image-2.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip gpt-image-2.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Run the mode check first so the agent knows whether it can generate images directly or only write prompts.
  • Keep API keys in a .env or ~/.gateway.env file rather than in prompts, and review scripts before enabling Garden mode since they make network calls.
  • Give specific details (subject, size, style, text) when asked; the templates ask targeted questions for missing fields.

What's inside

  • manifest.json
  • README.md
  • README.zh-CN.md
  • references/academic-figures/graphical-abstract.md
  • references/academic-figures/mechanism-diagram.md
  • references/academic-figures/method-pipeline-overview.md
  • references/academic-figures/multi-condition-comparison.md
  • references/academic-figures/neural-network-architecture.md
  • references/academic-figures/publication-chart.md
  • references/academic-figures/qualitative-comparison-grid.md
  • references/academic-figures/research-overview-poster.md
  • references/academic-figures/scientific-schematic.md
  • references/assets-and-props/game-screenshot-mockup.md
  • references/assets-and-props/retro-skeuomorphic-icons.md
  • SKILL.md

Redistributed unmodified from ConardLi/garden-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.