genimg

• One command for OpenAI and Google DeepMind image models, on an API key or your ChatGPT subscription through Codex.
• Variations that actually differ, side by side in an HTML grid, and a drawing studio to sketch what you mean.
• Skills so your coding agent can do all of it.

Made for humans and agents; the agent-native design is inspired by kenn-io's roborev and agentsview.

Docs: genimg.luischav.es

uv tool install genimg
genimg setup                      # connect a provider and save a profile
genimg "a minimal fox logo, NOT a grid" \
  --model oai:gi2.5-flare \
  --num 4 \
  --deltas "line art, block print, brush stroke" \
  --output fox.png \
  --grid \
  --open

Add the skills to Claude Code, Codex, Cursor or OpenCode with npx skills add lucharo/genimg: genimg plus workflows for infographics, visual exploration, visual review and image-to-app. Pick some with --skill, e.g. npx skills add lucharo/genimg --skill genimg genimg-infographic.

Docs

• Getting started: install, providers, first image
• For agents: the skills and what an agent can read
• Guide: diverse images, input images, Codex
• Reference: CLI, models, config.toml

Contributing

uv sync, then uv run pytest and uv run ruff check .. See CONTRIBUTING.md and the maintainer docs.