ComfyUI Controller Pro

支持批量生成10-100个修仙视频和图片，集成LTX2多版本模型与自动化浏览器及工作流管理功能。

Security Scan

The package does what its name says (a ComfyUI batch controller) but contains a locally‑exposed web server, arbitrary-model download functionality, and workflow/file upload endpoints that could unintentionally expose your local ComfyUI instance or allow remote files to be written — proceed with caution and review before running.

Purpose & Capability

The name/description (ComfyUI controller, LTX2 support, batch video/image generation, browser automation) match the included files: many scripts for batch generation, workflow conversion, an HTTP/SocketIO proxy server, browser automation scripts, and model download utilities. The requested capabilities are consistent with the stated purpose.

Instruction Scope

SKILL.md instructs running scripts that (a) proxy and control a ComfyUI instance at 127.0.0.1:8188, (b) download model weights from arbitrary URLs via scripts/download_weights.py, (c) run Playwright/browser automation, and (d) launch a Flask/SocketIO service. The code reads/writes files in user home and /tmp paths and exposes endpoints for uploading workflows and viewing/downloading outputs. Those operations go beyond simple local orchestration because they enable writing arbitrary files and bridging the local ComfyUI API to an HTTP endpoint — this is functionally required for the skill but increases the attack surface and risk of unintended data exposure.

Install Mechanism

There is no formal install spec in registry metadata, but SKILL.md and requirements.txt instruct pip installing playwright and requests. Playwright will additionally download browser binaries when installed/used. The repository itself includes many scripts and shell files; there is no hidden installer URL, which lowers risk, but the download_weights facility (and any scripts that fetch external model files) will fetch arbitrary remote content at runtime — intentional for the skill but a potential vector for malicious or very large payloads.

Credentials

The skill does not request environment variables or credentials in its metadata. That is proportionate: the code interacts with a local ComfyUI HTTP API, filesystem, and browser automation and does not require cloud credentials. One minor note: the code assumes ComfyUI is installed at ~/ComfyUI and hardcodes some paths and a Flask SECRET_KEY in server_enhanced.py (secret is static), which is not ideal but not itself a credential request.

Persistence & Privilege

The skill provides server_enhanced.py which runs a Flask/SocketIO service bound to 0.0.0.0:5005 with CORS '*' and proxies requests to the local ComfyUI API. That effectively exposes the local ComfyUI control surface over the network if run on a machine reachable by others. Endpoints allow workflow upload and trigger execution without authentication. always:false mitigates automatic global inclusion, but running the server as provided grants significant network-facing privileges and could be abused if exposed. The files also write to user dirs and /tmp (normal for such tooling) but should be run with caution.

What to consider before installing

This package is coherent with its advertised purpose, but review and take precautions before running: - Inspect scripts/download_weights.py and any download URLs before using the model-downloader — it will fetch arbitrary remote files into your ComfyUI folders. - Avoid running server_enhanced.py unmodified on a machine with public network access. By default it binds to 0.0.0.0 and enables CORS '*', which will expose your local ComfyUI API and allow anyone with network access to queue jobs or download outputs. Change the bind address to 127.0.0.1, restrict CORS/origins, and add authentication if you need remote access. - If you must test, run inside an isolated environment (VM, container, or on a host behind a firewall) and not on a machine holding sensitive data or credentials. - Review workflow upload handling (the API saves uploaded files to /tmp and passes them to WorkflowManager.upload_workflow) to ensure it cannot be used to overwrite important files; run as a non-privileged user. - Playwright will download browser binaries and can load remote content; make sure scripts that drive browsers don't visit untrusted URLs. - Check disk space and bandwidth: model files (safetensors/gguf) are large and downloads could exhaust disk or bandwidth. If you want, I can: (1) point to the specific lines in download_weights.py and server_enhanced.py that you should inspect and edit, (2) suggest minimal code changes to bind the server to localhost and disable CORS, or (3) produce a short checklist to safely run this package in a container.

SKILL.md - ComfyUI Controller

Skill Description

Advanced ComfyUI batch video and image generation system with LTX-2.3 support, browser automation, workflow execution, and model download capabilities.

Features

• 🎬 Batch Video Generation: Generate 10-100 videos in batch
• 🖼️ LTX2 Model Support: Multi-version LTX2 video generator (v1/v2/v3)
• 🌐 Browser Automation: Browser-based video generation
• ⚡ Workflow Execution: Run ComfyUI workflows via HTTP API
• ⬇️ Model Download: Auto-download model weights from URLs
• 🚀 Turbo Z Generation: Fast image generation
• 🎵 Batch Scene Generation: Music video multi-scene rendering
• 🔧 Workflow Management: Automatic workflow execution

Installation

pip install -r requirements.txt

System Requirements:

• Python 3.8+
• ComfyUI installed at ~/ComfyUI
• GPU ≥24GB VRAM (recommended for LTX2)
• playwright, requests

Usage

Batch Generate 10 Videos

python auto_generate_10_videos.py

Batch Generate 100 Xianxia News

python auto_run_100_xiuxian.py

LTX2 Video Generation (Recommended)

python generate_xiuxian_ltx2_v2.py

Recommended Settings:

• Steps: 15 (best balance, ~57 min)
• Frames: 72 (3s test) or 480 (20s final)
• Mode: I2V + LoRA

Run ComfyUI Workflow

# Edit workflow JSON first, then run
~/ComfyUI/venv/bin/python scripts/comfyui_run.py \
  --workflow assets/tmp-workflow.json

Download Model Weights

# Download from URL
echo "https://example.com/model.safetensors" | \
~/ComfyUI/venv/bin/python scripts/download_weights.py \
  --base ~/ComfyUI

# Specify subfolder
~/ComfyUI/venv/bin/python scripts/download_weights.py \
  --base ~/ComfyUI --subfolder loras \
  https://example.com/lora.safetensors

Browser Video Generation

python browser_video_gen.py

Turbo Z Image Generation

python z_turbo_generate.py

Batch Scene Generation (Music Video)

Use scripts/batch_scenes.js for automation:

// Run in browser console
const scenes = [
  {
    name: "scene_01",
    prompt: "A lonely girl running through rain at night",
    image: "unified_ref.png",
    steps: 15,
    frames: 72
  }
];

for (const scene of scenes) {
  await comfyui_batch.configureScene(scene);
  // Click Run button
}

File Structure

comfyui-controller/
├── auto_generate_10_videos.py      # Batch generate 10 videos
├── auto_run_100_xiuxian.py         # Auto run 100 xianxia news
├── batch_xiuxian_videos.py         # Batch process xianxia videos
├── browser_video_gen.py            # Browser video generator
├── generate_xiuxian_ltx2_v2.py     # LTX2 v2 generator
├── z_turbo_generate.py             # Turbo Z generator
├── comfyui_smart_controller.py     # Smart controller main
├── scripts/
│   ├── comfyui_run.py              # Workflow execution
│   ├── download_weights.py         # Model weight downloader
│   └── batch_scenes.js             # Batch scene automation
├── assets/
│   ├── default-workflow.json       # Default workflow
│   └── tmp-workflow.json           # Temporary workflow
└── LTX2_VIDEO_GUIDE.md             # LTX2 best practices

Performance Reference

Task	Speed	Description
10 videos	~5 min	Batch processing
100 news	~15 min	Batch processing
LTX2 video (15 steps)	~57 min/scene	I2V + LoRA
LTX2 video (25 steps)	~1h45m/scene	Final quality
Turbo Z image	~10 sec/image	Fast generation
Workflow execution	~30-60 sec	Single image

Best Practices

1. 15 steps is the sweet spot - I2V converges at 15-20 steps
2. Unified reference image - Same input for all scenes
3. Reload workflow every time - Avoid VAEDecode failures
4. Never reload during execution - Current run will fail
5. Frame selection - 72 frames (3s) for testing, 480 frames (20s) for final

Troubleshooting

Q: ComfyUI server not reachable? A: Check if ~/ComfyUI/main.py exists and server is running on 127.0.0.1:8188

Q: VAEDecode validation failed? A: Reload entire workflow, wait for models to fully load

Q: Output video not saved? A: Check for VAEDecode errors in log, then re-run

Q: How to download models? A: Use scripts/download_weights.py with URLs

License

MIT License

Author

yun520-1

GitHub

https://github.com/yun520-1/ComfyUI-Controller-Little-bug