Generate and refine images from text prompts with JSON metadata.
npx skills add https://github.com/LocalSymmetry/lofn --skill image-gen-localsymmetry
Please help me install this Agent Skill. Skill: image-gen Source: https://github.com/LocalSymmetry/lofn/tree/main/skills/image-gen Command: npx skills add https://github.com/LocalSymmetry/lofn --skill image-gen-localsymmetry
This Skill removes the friction of creating, editing, and refining images across multiple providers, so you can move from a concept to a finished visual asset without stitching together separate tools.
Use the image-gen skill to generate a 9:16 solarpunk garden city image and save it as ./images/garden.png.
High-intent search queries and answers about installing and using this skill.
To generate and refine images from text prompts, use a pipeline that supports multi-provider generation, image editing, aspect-ratio control, and JSON metadata output. This enables iterative corrections for visual details like hands and faces while preserving style.
Yes, you can fix anatomy, faces, clothing, and consistency in generated images without altering the style. Image editing and refinement models like Gemini nano-banana 2 or Flux Kontext handle these fast corrections while preserving the original composition.
The best way to automate a prompt-to-image generation workflow is to use a pipeline with built-in fallback behavior and JSON sidecar metadata output. This ensures reproducible outputs and continuous generation even if a primary provider fails.
Yes, this approach works well for portrait and social media posters. It applies to portrait-oriented outputs and social media workflows, using FAL Flux Pro 1.1 Ultra for high-quality compositions tailored to specific aspect ratios like 9:16.
Limitations of using Flux and Gemini for image editing include dependency on multi-provider availability and the need for fallback handling. Complex iterative edits may require multiple correction cycles to achieve desired visual consistency and detail accuracy.
HuaiminHuangGenerate images via the MiniMax API from text prompts and reference images.
FatOtterExtract JPEG thumbnails and metadata from media files using ffprobe and Pillow.
product-on-purposeInitialize JPKB projects with standardized documentation and folder structure.
bestGenerate and edit images via the Gemini Image Generation API with provider fallback.
