Skip to main content
Glama
deenesjakoruzh
yt0912iii

Independent Video Generator

by yt0912iii
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Hybrid

Independent Video Generator

This is a standalone video generation frontend and backend extracted from the main project. It only keeps the ComfyUI video workflow and does not include Agent conversations or image generation.

Startup

  1. Copy .env.example to .env and verify the ComfyUI and FFPROBE paths.

  2. Install dependencies: npm install

  3. Development mode: npm run dev

  4. Production build: npm run build

  5. Production start: npm run start

Default URL: http://127.0.0.1:8834

Windows users can simply double-click start.bat; it will automatically switch to the current folder, create .env, install dependencies, and start the service.

Related MCP server: RunComfy MCP Server

Video API

  • GET /api/video/workflows

  • POST /api/video/jobs

  • GET /api/video/jobs/:jobId

  • DELETE /api/video/jobs/:jobId

MCP Minimal Version

Start the video API first, then run npm run mcp. The MCP server uses stdio transport and calls the existing video API through MCP_API_URL (default http://127.0.0.1:8834).

Provided tools:

  • list_video_workflows

  • create_video_job

  • upload_video_asset

  • get_video_job

  • list_video_jobs

  • get_video_result

  • cancel_video_job

Remote frontend and GPU host

Keep the frontend and MCP on the portable control side, and keep Backend, ComfyUI, models, and GPU on the same GPU host. Build the frontend with:

VITE_BACKEND_URL=http://GPU_HOST:8834

Configure the GPU host Backend with:

HOST=0.0.0.0
PORT=8834
COMFYUI_URL=http://127.0.0.1:8188
CORS_ORIGIN=http://FRONTEND_HOST

Configure MCP with:

MCP_API_URL=http://GPU_HOST:8834

The browser and MCP call only the Video Generator API; ComfyUI remains private on the GPU host. API key authentication is not enabled yet, so HOST=0.0.0.0 should only be used on a trusted LAN or VPN.

Currently, images and audio are still transmitted as data URLs, which is suitable for a local MVP but not for large or multi-user deployments.

Video workflow assets are located in workflow/. This project currently shares the same ComfyUI service as the main project, but the code, frontend pages, job API, and workflow JSON are independent.

This server cannot be deployed

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

  • RunComfy MCP Server

    avalla
    F
    license
    Not graded
    quality
    D
    maintenance
    Enables generation of AI videos and images using RunComfy APIs. Supports multiple models for text-to-video, image-to-video, text-to-image, and image-to-image workflows with customizable parameters like aspect ratio, duration, and seed.
    -

  • ComfyUI MCP Server

    yarnovo
    A
    license
    Not graded
    quality
    D
    maintenance
    Dynamically loads ComfyUI workflows as MCP tools, enabling AI assistants to generate images, videos, and audio by executing workflows across categories like text-to-image, image-to-video, and text-to-audio with automatic parameter mapping and progress monitoring.
    43,607 npm
    3
    MIT

  • mcp-comfy-ui-builder

    MIt9
    A
    license
    Not graded
    quality
    D
    maintenance
    Enables discovery of ComfyUI nodes and building/managing workflows with real-time execution via WebSocket. Provides 50+ tools for node discovery, workflow building, template usage, model management, and batch/chain execution.
    89 npm
    3
    MIT