← 返回 Skills 市场

ComfyUI Generator

Name: ComfyUI Generator
Author: nl108

作者 nl108 · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ⚠ suspicious

114

总下载

当前安装

版本数

在 OpenClaw 中安装

/install comfyui-generator

功能描述

Generate AI images and perform style transfers via ComfyUI with batch processing and automated workflow management through OpenClaw integration.

使用说明 (SKILL.md)

ComfyUI Generator Skill

Overview

Integration between OpenClaw and ComfyUI for AI media generation.

Features

Text-to-image generation
Image style transfer
Batch processing
Automated workflow management
File monitoring and archiving

Installation

Prerequisites

ComfyUI installed at C:\ComfyUI
ComfyUI service running on http://127.0.0.1:8188
OpenClaw 2026.4.5+

Setup

# Set environment variables
export COMFY_API_KEY="local_comfyui"
export COMFY_BASE_URL="http://127.0.0.1:8188"

# Copy workflow files
cp -r workflows/*.json "C:\ComfyUI\input\workflows\"

Usage

Generate Image

openclaw comfy generate --prompt "a beautiful landscape" --style cyberpunk

Style Transfer

openclaw comfy style --image "path/to/image.jpg" --style "van gogh"

Batch Processing

openclaw comfy batch --input prompts.txt --output output_dir --iterations 5

Configuration

Environment Variables

COMFY_API_KEY="local_comfyui"
COMFY_BASE_URL="http://127.0.0.1:8188"
COMFY_OUTPUT_DIR="C:\ComfyUI\output"

Workflow Configuration

Place workflow JSON files in:

C:\ComfyUI\input\workflows\

Examples

Basic Image Generation

from comfy_client import ComfyUIClient

client = ComfyUIClient()
result = client.generate_image(
    prompt="a futuristic city at night",
    workflow="image_generation.json"
)

Style Transfer

result = client.style_transfer(
    image_path="input.jpg",
    style="cyberpunk",
    workflow="style_transfer.json"
)

Troubleshooting

Service Not Running

# Check if ComfyUI is running
curl http://127.0.0.1:8188

# Start ComfyUI
python "C:\ComfyUI\main.py" --listen 127.0.0.1 --port 8188

API Connection Issues

Verify firewall allows port 8188
Check ComfyUI logs in C:\ComfyUI\logs\
Ensure API key is set correctly

Generation Quality

Use detailed prompts
Adjust workflow parameters
Try different models

Files

Core Scripts

scripts/comfy_client.py - API client
scripts/prompt_generator.py - Prompt optimization
scripts/file_monitor.py - Output monitoring

Workflow Files

workflows/image_generation.json - Basic image generation
workflows/style_transfer.json - Style transfer
workflows/upscale.json - Image upscaling

Configuration

config/settings.yaml - Skill configuration
config/prompt_templates.json - Prompt templates

License

MIT License

Support

For issues and questions, please check:

ComfyUI documentation
OpenClaw documentation
Skill logs in C:\Users\LEI\.openclaw\logs\

安全使用建议

This skill appears to implement a legitimate ComfyUI client and a prompt generator, but there are notable inconsistencies and missing pieces. Before installing or running it: - Verify provenance: the skill has no homepage and unknown source. Prefer skills with a clear repository and maintainer. - Inspect and obtain missing files: SKILL.md references scripts/file_monitor.py and several workflow/config files that are not included. Ask the publisher for the missing files or a full repo before trusting the package. - Confirm environment variables: the package uses COMFY_BASE_URL, COMFY_API_KEY, and COMFY_OUTPUT_DIR but the registry metadata doesn't declare them. Treat any API key as sensitive; only set it if you understand its scope and origin. - Avoid running as admin: when testing, run in an isolated VM or sandbox and do not run copy commands that write into system or other app directories unless you trust the source. - Review network behavior: comfy_client.py appears to call only the configured base_url (default localhost). Ensure COMFY_BASE_URL is pointed at a local ComfyUI instance and not an external server you don't control. - If you need this skill, request a full source repo or manifest that includes the missing workflow and monitor scripts and ask the author why the metadata omits the used env vars and referenced files. If you cannot obtain satisfactory answers, do not install on sensitive systems.

能力评估

⚠ Purpose & Capability

Name/description (ComfyUI image generation, style transfer, batch processing) align with the included client and prompt-generator code. However SKILL.md and the README reference additional scripts (scripts/file_monitor.py), workflow and config files (workflows/*.json, config/*.yaml), and runtime behavior (file monitoring/archiving) that are not present in the package manifest. Also SKILL.md instructs copying workflows into C:\ComfyUI, which requires filesystem access to an external app directory not managed by the skill. These missing artifacts and external path operations are inconsistent with the declared file list and raise questions about completeness and intent.

⚠ Instruction Scope

Runtime instructions target a local ComfyUI service (http://127.0.0.1:8188) and include commands to set COMFY_* env vars, copy workflow files into C:\ComfyUI, check the service with curl, and start ComfyUI via python. The instructions are otherwise scoped to local ComfyUI endpoints (no external network endpoints). However SKILL.md mentions file_monitor and several config/workflow files that are missing; that implies the skill expects to read/monitor and archive files in user directories — a scope expansion not supported by the provided code. The instructions also reference user-specific log paths (C:\Users\LEI\.openclaw\logs\) which is a narrow privacy concern and suggests the author tested on a specific machine.

✓ Install Mechanism

There is no install spec (instruction-only install). No downloads, package registry installs, or archive extracts are requested by the manifest. That keeps disk-write/remote-install risk low. The package includes Python scripts which an agent could execute, so normal caution about running third-party scripts applies.

⚠ Credentials

Registry metadata lists no required environment variables, but both SKILL.md and comfy_client.py use COMFY_BASE_URL, COMFY_API_KEY and suggest COMFY_OUTPUT_DIR. The client will add an Authorization: Bearer header if COMFY_API_KEY is set. Asking for an API key is reasonable for an API client, but the manifest's omission of these env vars is an inconsistency. No other unrelated credentials are requested.

✓ Persistence & Privilege

The skill does not request always:true, does not claim to modify other skills or system-wide settings, and has no install steps that persist a background service. It does instruct copying workflows into an external ComfyUI directory, which writes files into another application's area, but that is a functional requirement rather than a persistence-privilege escalation.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install comfyui-generator
安装完成后，直接呼叫该 Skill 的名称或使用 /comfyui-generator 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.0

- Initial release of the ComfyUI Generator Skill. - Integrates OpenClaw with ComfyUI for AI media generation. - Supports text-to-image, style transfer, batch processing, and automated archiving. - Provides environment variable and workflow file setup instructions. - Includes example OpenClaw commands and Python usage samples. - Adds troubleshooting and configuration guidance.

元数据

Slug comfyui-generator

版本 1.0.0

许可证 MIT-0

累计安装 1

当前安装数 1

历史版本数 1

常见问题