image-generation
Installation
SKILL.md
Image Generation Orchestrator
Universal skill for generating images using available MCP tools. Handles tool selection, prompt engineering, character consistency, and batch workflows.
Announce: "Using image-generation skill to [generate/orchestrate] visuals with [tool name]."
Prerequisites
Before using this skill, check which tools are available. Not all are required — use what you have.
| Tool | MCP Server | API Key | How to get |
|---|---|---|---|
| Ideogram 3.0 | ideogram — python3 <path>/ideogram-mcp/src/ideogram_mcp/server.py |
IDEOGRAM_API_KEY |
https://ideogram.ai/manage-api — create key in dashboard |
| Luma Dream Machine | luma — python3 <path>/luma-mcp/src/luma_mcp/server.py |
LUMAAI_API_KEY |
https://lumalabs.ai/api — generate API key |
| Nanobanana | Gemini extension (built into Claude Code) | Gemini API access | Pre-configured if Gemini is set up |
Setup: Add the MCP server to your project's .mcp.json, set the API key in your environment or .env, restart Claude Code.
At skill start: Check which tools are actually available. If a tool is missing, say which one and why (e.g., "Ideogram not available — ideogram_generate tool not found. Install the Ideogram MCP server and set IDEOGRAM_API_KEY."). Continue with whatever tools ARE available.