Generate AI images via authenticated Codex, Pollinations, and Gemini providers.
npx skills add https://github.com/gracefullight/krds --skill oma-image-gracefullight
Please help me install this Agent Skill. Skill: oma-image Source: https://github.com/gracefullight/krds/tree/main/.agents/skills/oma-image Command: npx skills add https://github.com/gracefullight/krds --skill oma-image-gracefullight
It solves the problem of generating high-quality AI images reliably across different vendors while handling authentication, references, costs, and reproducible outputs.
manifest.json capturing prompt, vendor/model, and strategy attempts.Generate an illustration by running oma image generate with your prompt, using automatic vendor selection and default settings.
High-intent search queries and answers about installing and using this skill.
Multi-vendor image generation routes a single prompt through authenticated providers like Codex, Pollinations, and optionally Gemini, automatically dispatching to available vendors and writing a reproducibility manifest with the output images.
Reference-image support accepts explicit image paths and forwards them as structured visual references to supported vendors, ensuring the original images are used directly for guidance rather than being re-described as text prompts.
Prompt amplification satisfies prompt clarification requirements by refining and expanding the initial input before dispatching it to provider APIs, ensuring higher quality illustrations, product photos, and AI art outputs.
Multi-vendor routing dispatches the same prompt to multiple authenticated providers with strict comparison behavior, allowing you to A/B test outputs from Codex, Pollinations, and Gemini while controlling sizing, quality, and count limits.
Cost and safety guardrails enforce input validation, require cost confirmation at or above a specified threshold, and return clear exit codes for safety, authentication, invalid input, and timeout scenarios during image generation.
The generation process writes output images alongside a manifest.json file that captures the original prompt, selected vendor and model, and strategy attempts, ensuring full reproducibility for visual asset creation.
zendev-shIntegrate AI models and services into Go applications via a unified SDK.
citedyOptimize app store metadata, visuals, and localization for iOS and Android listings.
shenxingyGenerate and edit images via prompt-engineered Gemini Nano Banana requests.
first-flukeGenerate images from text prompts using multi-vendor AI models.