gargantuax/openskills
2 skills26.8k installs4 starsGitHub
Install
npx skills add https://github.com/gargantuax/openskillsSkills in this repo
1Gpt Image 2GPT Image 2 is an agent skill that wraps the full OpenAI-compatible GPT Image 2 surface in one Python 3.10+ script for skill-aware agents and manual runs. Solo builders use it when they need dependable text-to-image batches, mask-based edits, multi-reference edits, or mixed text-and-image prompts via the Responses API without stitching three different clients. The skill enforces GPT Image 2 constraint checks before network calls, supports streaming with partial image previews, and writes predictable output files with automatic de-duplication. Configuration flows from flags, environment variables, or .env with a documented precedence order, which fits indie workflows where API keys live in local env files. It complements design iteration during frontend builds while also feeding launch distribution assets and growth experiments that need fresh visuals on demand.16.6kinstalls2NanobananaNanobanana is an agent skill that wraps Gemini-native Nano Banana image generation and editing in the same pattern as repository skills like gpt-image-2: a single Python CLI, offline validation, `.env` support, and dry runs agents can sanity-check before spending quota. Solo builders use it when they need hero images, UI mocks, ad creatives, or batch variants without leaving Claude Code, Cursor, or Codex. You point it at `gemini-2.5-flash-image`, `gemini-3.1-flash-image-preview`, or `gemini-3-pro-image-preview` (or aliases nanobanana, nanobanana-2, nanobanana-pro), supply `GEMINI_API_KEY`, and run generate or batch with optional custom base URLs for gateways or self-hosted Gemini-compatible endpoints. Pre-flight checks catch unsupported aspect ratios and image sizes per model so failed requests do not burn agent turns. It fits indie SaaS and content shops that want repeatable image pipelines in repo scripts rather than manual playground copy-paste.10.2kinstalls