Requirements
- Target platform
- OpenClaw
- Install method
- Manual import
- Extraction
- Extract archive
- Prerequisites
- OpenClaw
- Primary doc
- SKILL.md
本地 ComfyUI 画图工作流 + CivitAI 集成。通过 API 控制本地 ComfyUI 生成图片(文生图/图生视频),支持 CivitAI 模型搜索/下载/更新检查/自动调参。Use when: 用户说画图、生成图片、generate image、comfyui、换模型、调参数、出图、搜模型、下载模型...
本地 ComfyUI 画图工作流 + CivitAI 集成。通过 API 控制本地 ComfyUI 生成图片(文生图/图生视频),支持 CivitAI 模型搜索/下载/更新检查/自动调参。Use when: 用户说画图、生成图片、generate image、comfyui、换模型、调参数、出图、搜模型、下载模型...
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete.
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run.
通过 ComfyUI API 在本地 RTX 5090 上生成图片,集成 CivitAI 模型管理。
检查 ComfyUI 是否运行,未运行则自动启动 解析用户需求:提示词、模型、参数 自动调参:根据模型从 CivitAI 获取推荐参数(或使用 config 缓存) 调用 generate.py 生成图片 将图片发送到 Discord 心跳时调用 auto_shutdown.py 检查空闲关闭
所有脚本相对于此 skill 目录: scripts/comfyui_manager.py — 启动/关闭/状态 scripts/generate.py — 生成图片(Python API 调用) scripts/auto_shutdown.py — 空闲自动关闭检查 scripts/civitai.py — CivitAI 集成(搜索/详情/下载/更新检查/推荐参数) scripts/auto_tune.py — 自动调参(从 CivitAI 样图元数据提取推荐设置) config.json — 配置文件(模型别名、默认参数、路径、CivitAI 关联)
import scripts.comfyui_manager as mgr mgr.start() # 自动从 WSL 启动 Windows ComfyUI 进程
from scripts.generate import generate result = generate( positive="your prompt here", model="noobv6", # 见下方模型别名 batch_size=4, width=1024, height=1536, steps=20, cfg=5, ) # result = {"ok": True, "images": ["/mnt/d/.../sdi_00201_.png", ...]}
from scripts.auto_tune import get_tuned_params from scripts.generate import generate params = get_tuned_params("janku") # 从 CivitAI 获取推荐参数 result = generate( positive="your prompt", model="janku", steps=params["steps"], cfg=params["cfg"], width=params["width"], height=params["height"], )
python3 scripts/civitai.py search "anime checkpoint" --limit 5
python3 scripts/civitai.py info 140272
python3 scripts/civitai.py params 140272
python3 scripts/civitai.py check-updates ⚠️ 需要计算 SHA256,每个文件约 30 秒
python3 scripts/civitai.py download "https://civitai.com/api/download/models/XXXXX" --filename model_name.safetensors
python3 scripts/auto_tune.py update-all
生成完成后,将图片 cp 到 workspace 临时目录,用 message tool 发送到 #🎨-画图 频道(channel:1476675131404193823)。
import scripts.comfyui_manager as mgr mgr.stop()
别名CheckpointCivitAI ID版本风格hassakuhassakuXLIllustrious_v34140272v3.4Illustrious 动漫noobv6pornmasterPro_noobV61045588noob-V6NoobAI 动漫(默认)noobv4pornmasterPro_noobV41045588noob-V4NoobAI 动漫(旧版)sdxlv8pornmaster_proSDXLV882543Pro-SDXL-V8SDXL 写实jankuJANKUTrainedNoobaiRouwei_v691277670v6.9NoobAI+RouWei 混合nova3dnova3DCGXL_ilV80715287IL v8.03DCG/2.5DunholyunholyDesireMixSinister_v701307857v7.0暗黑美学fluxflux1-dev-bnb-nf4-v2638187BNB NF4 v2Flux 写实
每个模型在 config.json 中有 CivitAI 推荐的最佳参数。生成时: 用户指定的参数优先 未指定则使用 config.json → model_params[alias] 的推荐值 model_params 也没有则使用 defaults 全局默认 模型StepsCFG采样器hassaku207euler_ajanku207euler_anoobv6207euler_asdxlv8255dpmpp_2mnova3d207euler_aunholy207euler_aflux201euler
API Key 存于 ~/.openclaw/workspace/credentials/civitai.md,格式: # CivitAI API Key Token: your_api_key_here
用户给自然语言描述时,转换为专业 Stable Diffusion 提示词: 质量标签在前:masterpiece, best quality, absurdres, highres 用 BREAK 分隔不同语义段 动作/姿势用专业标签:mating press, pov, from below, spread legs 细节标签:detailed skin, sweat, wet skin, motion blur, dynamic angle 负面提示词加入:censored, mosaic censoring, bar censor
每次心跳时运行 auto_shutdown.py 检查空闲时间,超过 15 分钟无使用自动关闭 ComfyUI 释放显存。
Code helpers, APIs, CLIs, browser automation, testing, and developer operations.
Largest current source with strong distribution and engagement signals.