Skip to content
Get packs

AI Image Generation

Lets your agent generate images via several image-generation APIs and batch-produce them from prompt files

Source Download skill
Repository
JimLiu/baoyu-skills
Stars
27,191
Licence
MIT
Files
15 files

What it does

Once installed, the agent can create images from text descriptions using providers such as OpenAI, Google, OpenRouter, DashScope, Replicate and others, including using reference photos to keep a person or object consistent. It asks you to choose a provider, model and quality the first time, saves those preferences, and can generate many images at once from a batch file.

Great for

  • Creators who want illustrations for articles or posts generated from written prompts
  • Designers who want to iterate on images with reference pictures and consistent subjects
  • Users who already have a set of prompt files and want them rendered in one batch

Skip it if

  • Users without an API key or a logged-in Codex subscription, since image generation requires paid provider access (the skill cannot run without one)
  • Anyone who needs to use real people's likenesses without consent, since the skill is built to preserve a referenced person's identity

Try asking

  • “Generate a 16:9 illustration of a mountain lake at sunrise and save it as lake.png”
  • “Make the product in my reference photo blue, keeping everything else the same”
  • “Build a batch file from my outline.md and prompts folder and generate all the images with 4 workers”
Step-by-step guide: how to use AI Image GenerationInstall, 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.

    baoyu-image-gen.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip baoyu-image-gen.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip baoyu-image-gen.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

  • Give the agent 2-4 clean reference images rather than many large ones when preserving a subject's identity
  • Keep identity-preservation prompts short and state that references show the same subject, instead of long facial descriptions
  • Check the EXTEND.md preferences file at the project or home path so you know which provider and model will be charged

What's inside

  • references/codex-image2-fallback.md
  • references/codex-oauth-vs-openai-api-key.md
  • references/config/first-time-setup.md
  • references/config/preferences-schema.md
  • references/providers/agnes.md
  • references/providers/codex-cli.md
  • references/providers/dashscope.md
  • references/providers/minimax.md
  • references/providers/openrouter.md
  • references/providers/replicate.md
  • references/providers/zai.md
  • references/usage-examples.md
  • scripts/build-batch.test.ts
  • scripts/build-batch.ts
  • SKILL.md

Redistributed unmodified from JimLiu/baoyu-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.