@openpets/imagegen
v1.0.0
Published
Multi-provider image and video generation plugin. Aggregates tools from FAL (Flux, Stable Diffusion, Kling Video), OpenAI (DALL-E, GPT Image), Google Gemini, and HuggingFace into a single interface with task-oriented presets for logos, posters, banners, a
Maintainers
Readme
ImageGen
Unified image and video generation across multiple providers. Aggregates tools from FAL, OpenAI, and Google Gemini into a single plugin with task-oriented presets.
Providers
| Provider | Image Gen | Image Edit | Video Gen | Image Animate | |----------|-----------|------------|-----------|---------------| | FAL | Flux, Stable Diffusion | Style transfer, editing | Kling, Runway Gen-3 | Image-to-video | | OpenAI | DALL-E, GPT Image | GPT Image edit | - | - | | Google Gemini | Prompt enhancement | Placeholder (Imagen 3) | Placeholder | - |
Setup
Set at least one provider's API key:
# FAL AI (recommended for image + video)
FAL_KEY=your-fal-key
# OpenAI (DALL-E / GPT Image)
OPENAI_API_KEY=your-openai-key
# Google Gemini
GOOGLE_API_KEY=your-google-key
# Optional: bypass provider-level read-only filtering for imported tools
IMAGEGEN_BYPASS_READONLY=trueWhen IMAGEGEN_BYPASS_READONLY=true, ImageGen loads provider tools without applying the provider pet's own read-only filter (for example OpenAI image/video write tools). ImageGen still respects its own read-only setting for ImageGen-native composite tools.
Dimension Presets
The imagegen-create tool auto-selects dimensions per provider:
| Preset | Description | Aspect |
|--------|-------------|--------|
| logo | Square logo/icon | 1:1 |
| poster | Vertical poster | 3:4 |
| banner | Wide horizontal banner | 16:9 |
| social | Social media post | 1:1 |
| story | Vertical story/reel | 9:16 |
| thumbnail | YouTube/video thumbnail | 16:9 |
| cover | Album/book cover | 1:1 |
| wallpaper | Desktop wallpaper | 16:9 |
| portrait | Portrait photo | 3:4 |
| landscape | Landscape photo | 4:3 |
Usage
# List available providers and tools
opencode run "list all available image generation providers"
# Generate with preset dimensions
opencode run "create a logo for a coffee shop called Bean There"
opencode run "create a poster for a summer music festival"
opencode run "create a banner for my tech blog"
# Generate video
opencode run "generate a video of ocean waves at sunset"
# Use specific provider
opencode run "generate a portrait using openai"Tools
ImageGen Tools (composite)
imagegen-list-providers- List all configured providers and their toolsimagegen-create- Generate image with preset dimensions, auto-routes to best providerimagegen-create-video- Generate video, auto-routes to best provider
FAL Tools (when FAL_KEY is set)
fal-generate-image- Text-to-image (Flux, Stable Diffusion)fal-transform-image- Image-to-image (style transfer, editing)fal-generate-video- Text-to-video (Kling, Runway Gen-3)fal-animate-image- Image-to-video animation
OpenAI Tools (when OPENAI_API_KEY is set)
openai-create-image- DALL-E / GPT Image generationopenai-create-image-edit- Image editingopenai-create-image-variation- Image variations (DALL-E 2)
Google Gemini Tools (when GOOGLE_API_KEY is set)
gemini-text-to-image- Prompt enhancement for image generationgemini-image-to-image- Placeholder for future Imagen 3gemini-text-to-video- Placeholder for future video generation
