jau123
MeiGen-AI-Design-MCP
Open-source MCP server for AI image & video generation, supporting 11 models, 1,446 curated prompts, and parallel sub-agent orchestration.
Repository last updated Jun 23, 2026
Overview
MeiGen AI Design MCP is an open-source MCP server that transforms AI coding tools into professional design assistants. It connects to multiple backends—MeiGen cloud, OpenAI-compatible APIs, or local ComfyUI—to generate images and videos. The server includes a library of over 1,400 curated prompts and supports parallel batch generation via specialized sub-agents.
With support for 11 leading models including Midjourney, Flux, and Seedance, MeiGen works seamlessly with Claude Code, Cursor, Windsurf, Hermes Agent, and any MCP-compatible host. It also offers a standalone CLI mode for shell scripts and CI pipelines.
This plugin is ideal for developers and designers who want to integrate AI-powered visual content generation directly into their coding workflow, whether for rapid prototyping, batch production, or creative exploration.
Highlights
- Supports 11 models including Midjourney, Flux, and local ComfyUI
- Includes 1,446 curated prompt templates and style enhancement
- Features parallel sub-agent orchestration for batch generation
- Works with Claude Code, Cursor, Windsurf, Hermes Agent, and more
- Offers standalone CLI mode for scripts and CI pipelines
Features
Multi-Backend Support
Choose between MeiGen cloud, any OpenAI-compatible API, or local ComfyUI for image and video generation.
Curated Prompt Library
Access 1,446 trending prompt templates with visual previews and style-aware enhancement.
Parallel Sub-Agent Orchestration
Specialized agents handle prompt crafting, image generation, and gallery research in parallel, keeping the main context clean.
Slash Commands
Use /meigen:gen, /meigen:find, /meigen:models, and /meigen:setup for quick actions directly in the chat.
Standalone CLI Mode
Run npx meigen gen for one-shot generation without an MCP host, ideal for automation.
Automation Hooks
Config validation on session start and auto-open of generated images on macOS.
Installation
Commands are reproduced exactly as published by the project maintainers. Always check the upstream repository for the current instructions.
# Add the plugin marketplace
/plugin marketplace add jau123/MeiGen-AI-Design-MCP
# Install
/plugin install meigen@meigen-marketplace/plugin marketplace add wshobson/agents
/plugin install meigen-ai-design@claude-code-workflows/meigen:setupnpx meigen init cursor # Cursor
npx meigen init vscode # VS Code / GitHub Copilot
npx meigen init windsurf # Windsurf
npx meigen init roo # Roo Code
npx meigen init claude # Claude Code (project-level)Requirements
- Node.js (for npx usage)
- MCP-compatible host (e.g., Claude Code, Cursor, Windsurf, Hermes Agent) or shell for CLI mode
- API token for MeiGen cloud or OpenAI-compatible provider (optional for free features)
- ComfyUI instance (optional, for local generation)
How it's built
MeiGen AI Design MCP is built with TypeScript and distributed as an npm package. The repository follows a standard MCP server structure with a plugin/ directory containing agent definitions, slash commands, and hooks. The bin/meigen-mcp.js entry point launches the MCP server, which communicates via the Model Context Protocol.
The server integrates with three provider types: MeiGen cloud (via API token), OpenAI-compatible APIs (via key and base URL), and local ComfyUI (via URL). Configuration is stored in ~/.config/meigen/config.json, and environment variables override file settings. The package includes a data/trending-prompts.json file with 1,446 curated prompts.
For Hermes Agent, add the server to ~/.hermes/config.yaml with increased timeout and connect_timeout values to accommodate video generation and initial npx download.
Source
This summary was written for Neura Market from the project's own repository. Install commands, identifiers and licence details are reproduced unchanged. The jau123/MeiGen-AI-Design-MCP repository is the authoritative source and the place to file issues or contribute.
Category: MCP Servers · Licensed under MIT