image-generationlisted

# Image Generation Skill You have access to Google Gemini image generation via the `generate_image` MCP tool. ## When to Use Generate images when the user: - Asks for an image, illustration, or visual asset - Needs a hero image, placeholder, icon, or background for a project - Asks to edit, modify, or iterate on an existing image - Mentions creating visual content of any kind ## How to Use Call the `generate_image` tool. The behaviour depends on the parameters: ### Text-to-Image (no input images) ```json { "prompt": "A modern dashboard UI with dark theme and blue accent colours", "aspectRatio": "16:9", "resolution": "2K", "filename": "dashboard-hero", "subfolder": "landing-page" } ``` ### Image Editing (with input images) ```json { "prompt": "Change the background to a sunset over water", "images": ["./src/assets/hero.png"], "aspectRatio": "16:9" } ``` ### Multi-Turn Refinement The tool returns a `sessionId` with every response. Pass it back to continue editing: ```json { "prompt": "Make the colours warmer and add more contrast", "sessionId": "session-1711929600000-a1b2c3" } ``` This preserves the conversation history so the model remembers what it generated. ## Output Organization - Use `filename` to give images meaningful names (e.g. `hero-banner` instead of `gemini-1711929600000-a1b2c3`) - Use `subfolder` to group related assets (e.g. `landing-page`, `blog-posts`) - Duplicate filenames are auto-versioned: `hero.png`, `hero-v2.png`, `hero-v3.pn