image-gen

name	image-gen
description	AI image generation with OpenAI, Google and DashScope APIs. Supports text-to-image, reference images, aspect ratios, and direct Cloudflare R2 upload. Sequential by default; parallel generation available on request. Use when user asks to generate, create, or draw images.

Image Generation

Generate images via OpenAI, Google Gemini, or DashScope APIs. R2 upload is the default workflow for persistent article/social images.

Scripts

SKILL_DIR/
  scripts/main.ts        — generate image (+ optional R2 upload)
  scripts/upload-r2.ts   — upload any existing file to R2
  scripts/r2.ts          — R2 upload module (AWS SigV4, no external deps)
  scripts/providers/     — google.ts / openai.ts / dashscope.ts
  scripts/types.ts       — shared types

Usage

SKILL_DIR="/path/to/image-gen"

# Basic generation
npx -y bun ${SKILL_DIR}/scripts/main.ts --prompt "A cat" --image cat.png

# With aspect ratio
npx -y bun ${SKILL_DIR}/scripts/main.ts --prompt "A landscape" --image out.png --ar 16:9

# From prompt file
npx -y bun ${SKILL_DIR}/scripts/main.ts --promptfiles prompt.md --image out.png

# Generate + upload to R2 (prints public URL)
npx -y bun ${SKILL_DIR}/scripts/main.ts --prompt "A cat" --image cat.png --r2

# Generate + upload with custom R2 key
npx -y bun ${SKILL_DIR}/scripts/main.ts --prompt "A cat" --image cat.png --r2 --r2-key articles/hero.png

# Upload existing file to R2
npx -y bun ${SKILL_DIR}/scripts/upload-r2.ts --file image.png
npx -y bun ${SKILL_DIR}/scripts/upload-r2.ts --file image.png --key custom/path.png

# With reference image
npx -y bun ${SKILL_DIR}/scripts/main.ts --prompt "Make it blue" --image out.png --ref source.png

# Force provider
npx -y bun ${SKILL_DIR}/scripts/main.ts --prompt "A cat" --image out.png --provider openai

Options

Option	Description
-p, --prompt <text>	Prompt text
--promptfiles <files...>	Read prompt from files
--image <path>	Output path (required)
--provider google|openai|dashscope	Force provider
-m, --model <id>	Model ID
--ar <ratio>	Aspect ratio: 16:9, 1:1, 4:3, 9:16, 2.35:1
--quality normal|2k	Default: 2k
--imageSize 1K|2K|4K	Google only
--ref <files...>	Reference images
--r2	Upload to R2 after generation, print public URL
--r2-key <name>	R2 object key (default: filename)
--json	JSON output

Environment Variables

Set in the vault-local .env.r2, ~/.skills/.env, or ~/.baoyu-skills/.env:

Variable	Description
GOOGLE_API_KEY / GEMINI_API_KEY	Google API key
OPENAI_API_KEY	OpenAI API key
DASHSCOPE_API_KEY	DashScope API key
R2_ACCESS_KEY_ID	Cloudflare R2 access key
R2_SECRET_ACCESS_KEY	Cloudflare R2 secret key
R2_ACCOUNT_ID	Cloudflare account ID
R2_BUCKET	R2 bucket name
R2_PUBLIC_BASE_URL	Public base URL (e.g. https://cdn.example.com)

Provider Selection

Auto-detects from available API keys. Priority: Google → OpenAI → DashScope. With --ref: Google first (Gemini multimodal), then OpenAI.

R2 Upload

Uses AWS SigV4 signing with built-in Node.js crypto — no external dependencies. With --r2, prints the R2 public URL instead of the local path. For article/social images, always use --r2 and insert the returned URL into Markdown; keep local files only as temporary cache.

Generation Mode

Sequential by default. For parallel batch generation, use multiple background subagents.

Default Workflow in This Vault

For any image that will be referenced by articles, tweets, covers, or long-lived notes:

1. Generate the image to a temporary or ignored local path.
2. Upload immediately with --r2 or scripts/upload-r2.ts.
3. Insert the returned R2_PUBLIC_BASE_URL URL into Markdown.
4. Do not rely on committed local image files as the canonical asset.