Image Generation Mcp

This handles image generation through Gemini's API with some nice guardrails built in. Instead of firing off whatever prompt you write, it enforces a structured format covering subject, style, lighting, mood, composition, and brand colors, then runs validation across multiple agents before generating anything. You get 2-4 variants per request with full metadata for tracking what worked. It supports three Gemini models depending on whether you need speed or quality. The Korean documentation suggests this came out of a real marketing workflow where brand consistency mattered more than just getting images fast. If you're tired of prompt roulette and need reproducible results, the validation layer is worth the extra orchestration overhead.

Install to Claude Code

npx -y skills add supercent-io/skills-template --skill image-generation-mcp --agent claude-code