[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-jau123--MeiGen-AI-Design-MCP":3,"tool-jau123--MeiGen-AI-Design-MCP":64},[4,17,26,40,48,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,2,"2026-04-03T11:11:01",[13,14,15],{"id":27,"name":28,"github_repo":29,"description_zh":30,"stars":31,"difficulty_score":23,"last_commit_at":32,"category_tags":33,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,34,35,36,15,37,38,13,39],"数据工具","视频","插件","其他","语言模型","音频",{"id":41,"name":42,"github_repo":43,"description_zh":44,"stars":45,"difficulty_score":10,"last_commit_at":46,"category_tags":47,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,38,37],{"id":49,"name":50,"github_repo":51,"description_zh":52,"stars":53,"difficulty_score":10,"last_commit_at":54,"category_tags":55,"status":16},519,"PaddleOCR","PaddlePaddle\u002FPaddleOCR","PaddleOCR 是一款基于百度飞桨框架开发的高性能开源光学字符识别工具包。它的核心能力是将图片、PDF 等文档中的文字提取出来，转换成计算机可读取的结构化数据，让机器真正“看懂”图文内容。\n\n面对海量纸质或电子文档，PaddleOCR 解决了人工录入效率低、数字化成本高的问题。尤其在人工智能领域，它扮演着连接图像与大型语言模型（LLM）的桥梁角色，能将视觉信息直接转化为文本输入，助力智能问答、文档分析等应用场景落地。\n\nPaddleOCR 适合开发者、算法研究人员以及有文档自动化需求的普通用户。其技术优势十分明显：不仅支持全球 100 多种语言的识别，还能在 Windows、Linux、macOS 等多个系统上运行，并灵活适配 CPU、GPU、NPU 等各类硬件。作为一个轻量级且社区活跃的开源项目，PaddleOCR 既能满足快速集成的需求，也能支撑前沿的视觉语言研究，是处理文字识别任务的理想选择。",74913,"2026-04-05T10:44:17",[38,14,13,37],{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":23,"last_commit_at":62,"category_tags":63,"status":16},2471,"tesseract","tesseract-ocr\u002Ftesseract","Tesseract 是一款历史悠久且备受推崇的开源光学字符识别（OCR）引擎，最初由惠普实验室开发，后由 Google 维护，目前由全球社区共同贡献。它的核心功能是将图片中的文字转化为可编辑、可搜索的文本数据，有效解决了从扫描件、照片或 PDF 文档中提取文字信息的难题，是数字化归档和信息自动化的重要基础工具。\n\n在技术层面，Tesseract 展现了强大的适应能力。从版本 4 开始，它引入了基于长短期记忆网络（LSTM）的神经网络 OCR 引擎，显著提升了行识别的准确率；同时，为了兼顾旧有需求，它依然支持传统的字符模式识别引擎。Tesseract 原生支持 UTF-8 编码，开箱即用即可识别超过 100 种语言，并兼容 PNG、JPEG、TIFF 等多种常见图像格式。输出方面，它灵活支持纯文本、hOCR、PDF、TSV 等多种格式，方便后续数据处理。\n\nTesseract 主要面向开发者、研究人员以及需要构建文档处理流程的企业用户。由于它本身是一个命令行工具和库（libtesseract），不包含图形用户界面（GUI），因此最适合具备一定编程能力的技术人员集成到自动化脚本或应用程序中",73286,"2026-04-03T01:56:45",[13,14],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":76,"owner_company":76,"owner_location":76,"owner_email":76,"owner_twitter":78,"owner_website":76,"owner_url":79,"languages":80,"stars":97,"forks":98,"last_commit_at":99,"license":100,"difficulty_score":10,"env_os":101,"env_gpu":101,"env_ram":101,"env_deps":102,"category_tags":105,"github_topics":106,"view_count":10,"oss_zip_url":76,"oss_zip_packed_at":76,"status":16,"created_at":117,"updated_at":118,"faqs":119,"releases":125},781,"jau123\u002FMeiGen-AI-Design-MCP","MeiGen-AI-Design-MCP","MeiGen-AI-Design-MCP — Turn Claude Code \u002F OpenClaw into your local Lovart. Local ComfyUI, 1,400+ prompt library, multi-direction parallel generation.","MeiGen-AI-Design-MCP 是一款专为大语言模型设计的开源 MCP 服务器插件，致力于将代码智能体转化为专业的本地设计助手。它有效解决了 AI 绘图依赖外部服务、提示词编写门槛高以及生成结果单一等痛点。\n\n通过集成本地 ComfyUI，MeiGen-AI-Design-MCP 实现了完全离线的图像生成，无需担心 API 费用或隐私泄露。其内置的 1,400 多条精选提示词库，配合精细化的提示词工程技术，能自动将自然需求转化为高质量的绘图指令。更值得一提的是，它支持多方向并行生成与子代理执行，能在短时间内输出多种创意方案，保持主上下文窗口的整洁。\n\nMeiGen-AI-Design-MCP 非常适合开发者、视觉设计师以及追求本地化部署的研究人员使用。无论是快速迭代产品概念，还是探索艺术风格，都能借助其高效的工作流显著提升创作效率。","\u003Ch1 align=\"center\">\n  MeiGen AI Design MCP \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fpunkpeye\u002Fawesome-mcp-servers\">\u003Cimg src=\"https:\u002F\u002Fawesome.re\u002Fmentioned-badge.svg\" alt=\"Mentioned in Awesome MCP Servers\">\u003C\u002Fa> \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwshobson\u002Fagents\u002Ftree\u002Fmain\u002Fplugins\u002Fmeigen-ai-design\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fwshobson%2Fagents-Featured-blue?style=flat&logo=github\" alt=\"Featured in wshobson\u002Fagents\">\u003C\u002Fa>\n\u003C\u002Fh1>\n\n\u003Cp align=\"center\">\n  \u003Cstrong>Turn Claude Code \u002F OpenClaw into your personal design assistant rivaling Lovart\u003C\u002Fstrong>\u003Cbr>\u003Csub>Local ComfyUI generation, API integration, 1,300+ pro prompt library, multi-direction parallel output\u003C\u002Fsub>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002Fmeigen\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fnpm\u002Fv\u002Fmeigen?style=flat-square&color=blue\" alt=\"npm version\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002Fmeigen\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fnpm\u002Fdm\u002Fmeigen?style=flat-square&color=green\" alt=\"npm downloads\">\u003C\u002Fa>\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FType-MCP_Server-blue?style=flat-square\" alt=\"MCP Server\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLocal-ComfyUI-green?style=flat-square\" alt=\"ComfyUI Support\">\n  \u003Ca href=\"LICENSE\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-lightgrey?style=flat-square\" alt=\"MIT\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FuX6rnersUx\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDiscord-Join-5865F2?style=flat-square&logo=discord&logoColor=white\" alt=\"Discord\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"#quick-start\">Quick Start\u003C\u002Fa> •\n  \u003Ca href=\"#see-it-in-action\">Demo\u003C\u002Fa> •\n  \u003Ca href=\"#features\">Features\u003C\u002Fa> •\n  \u003Ca href=\"#providers\">Providers\u003C\u002Fa> •\n  \u003Ca href=\"#slash-commands\">Commands\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Cstrong>English\u003C\u002Fstrong> | \u003Ca href=\"README.zh-CN.md\">中文\u003C\u002Fa>\n\u003C\u002Fp>\n\n---\n\n## What Is This?\n\nAn open-source MCP Server (installed via plugin marketplace) that gives LLMs creative and aesthetic capabilities through 8 tools and carefully designed skills, enabling them to handle complex design tasks. It teaches LLMs how to use various image generation models effectively, delivering professional results through reference images and multi-direction parallel workflows.\n\n- Works with local ComfyUI — no external API dependency; also easily integrates with any custom API\n- Built-in 1,500+ curated prompt templates from [nanobanana-trending-prompts](https:\u002F\u002Fgithub.com\u002Fjau123\u002Fnanobanana-trending-prompts) and fine-tuned prompt engineering techniques that turn requirements into concrete image generation tasks\n- Parallel batch generation and sub-agent execution to keep the main context window clean\n\n---\n\n## See It in Action\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fyoutu.be\u002FJQ3DZ1DXqvs\">\n    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_b53e554df66c.jpg\" alt=\"Watch Demo\" width=\"400\">\n  \u003C\u002Fa>\n  \u003Cbr>\n  \u003Cb>\u003Ca href=\"https:\u002F\u002Fyoutu.be\u002FJQ3DZ1DXqvs\">▶ Watch demo on YouTube\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fp>\n\n### Product Photo — 4 Directions in Parallel\n\n> *\"Create 4 product display images for this perfume, one of which should feature a model.\"*\n\n**Process** — AI uploads the reference image, crafts 4 distinct prompts, then generates all 4 in parallel:\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_8982805ea0f6.jpg\" alt=\"Parallel generation process\" width=\"700\">\n\u003C\u002Fp>\n\n**Result** — 4 creative directions delivered in under 2 minutes:\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_b3ec99310670.jpg\" alt=\"Generation results\" width=\"700\">\n\u003C\u002Fp>\n\n**Generated images:**\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_6d440009d1b5.jpg\" alt=\"Luxury still life\" width=\"24%\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_4a8eb084a9a9.jpg\" alt=\"Model campaign\" width=\"24%\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_11ccb9c61940.jpg\" alt=\"Nature botanicals\" width=\"24%\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_6961434888ec.jpg\" alt=\"Minimalist editorial\" width=\"24%\">\n\u003C\u002Fp>\n\n---\n\n## Quick Start\n\n### Claude Code Plugin (Recommended)\n\n```bash\n# Add the plugin marketplace\n\u002Fplugin marketplace add jau123\u002FMeiGen-AI-Design-MCP\n\n# Install\n\u002Fplugin install meigen@meigen-marketplace\n```\n\n**Restart Claude Code** after installation (close and reopen, or open a new terminal tab).\n\n**Alternative marketplace** — also available via [wshobson\u002Fagents](https:\u002F\u002Fgithub.com\u002Fwshobson\u002Fagents) (30k+ stars):\n\n```bash\n\u002Fplugin marketplace add wshobson\u002Fagents\n\u002Fplugin install meigen-ai-design@claude-code-workflows\n```\n\n> This marketplace doesn't bundle MCP server config. After installing, add to your project's `.mcp.json`:\n> ```json\n> { \"mcpServers\": { \"meigen\": { \"command\": \"npx\", \"args\": [\"-y\", \"meigen@1.2.6\"] } } }\n> ```\n\n#### First-Time Setup\n\nFree features work immediately after restart — try:\n\n> \"Search for some creative inspiration\"\n\nTo unlock image generation, run the setup wizard:\n\n```\n\u002Fmeigen:setup\n```\n\nThe wizard walks you through:\n1. **Choose a provider** — local ComfyUI, MeiGen Cloud, or any OpenAI-compatible API (bring your own key & endpoint)\n2. **Enter credentials** — ComfyUI URL, API token, or key\n3. **Done** — restart Claude Code once more, then start generating\n\n### Cursor \u002F VS Code \u002F Windsurf \u002F Roo Code\n\nOne command to set up MeiGen for any supported AI coding tool:\n\n```bash\nnpx meigen init cursor      # Cursor\nnpx meigen init vscode      # VS Code \u002F GitHub Copilot\nnpx meigen init windsurf    # Windsurf\nnpx meigen init roo         # Roo Code\nnpx meigen init claude      # Claude Code (project-level)\n```\n\nThis writes the correct MCP config file with the right format and path for your tool. If a config file already exists, MeiGen is merged in without overwriting your other servers.\n\n### OpenClaw\n\nInstall the skill from [ClawHub](https:\u002F\u002Fclawhub.ai\u002Fjau123\u002Fcreative-toolkit):\n\n```bash\nnpx clawhub@latest install creative-toolkit\n```\n\nOr install the full plugin (includes commands, skills, and MCP server):\n\n```bash\nopenclaw plugins install meigen-ai-design\n```\n\n### Other MCP-Compatible Hosts\n\nAdd to your MCP config (e.g. `.mcp.json`, `claude_desktop_config.json`):\n\n```json\n{\n  \"mcpServers\": {\n    \"meigen\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"meigen@latest\"],\n      \"env\": {\n        \"MEIGEN_API_TOKEN\": \"meigen_sk_...\"\n      }\n    }\n  }\n}\n```\n\n> Free features (inspiration search, prompt enhancement, model listing) work without any API key.\n\n---\n\n\u003Ch2 id=\"features\">Features\u003C\u002Fh2>\n\n### MCP Tools\n\n| Tool | Free | Description |\n|------|------|-------------|\n| `search_gallery` | Yes | Search 1,500+ curated trending prompts with visual previews (powered by [nanobanana-trending-prompts](https:\u002F\u002Fgithub.com\u002Fjau123\u002Fnanobanana-trending-prompts)) |\n| `get_inspiration` | Yes | Get full prompt, all images, and metadata for any gallery entry |\n| `enhance_prompt` | Yes | Transform a brief idea into a professional image prompt |\n| `list_models` | Yes | List available models across all configured providers |\n| `upload_reference_image` | Yes | Compress and upload a local image for use as a reference |\n| `comfyui_workflow` | Yes | Manage ComfyUI workflow templates: list, view, import, modify, delete |\n| `manage_preferences` | Yes | Remember your preferred style, aspect ratio, model, and favorite prompts |\n| `generate_image` | Key | Generate an image — routes to the best available provider automatically |\n\n### Slash Commands\n\n| Command | Description |\n|---------|-------------|\n| `\u002Fmeigen:gen \u003Cprompt>` | Quick generate — skip conversation, go straight to image |\n| `\u002Fmeigen:find \u003Ckeywords>` | Search 1,300+ curated prompts for inspiration |\n| `\u002Fmeigen:models` | Browse and switch AI models for this session |\n| `\u002Fmeigen:ref \u003Cfile>` | Upload a local image as reference, get a URL back |\n| `\u002Fmeigen:setup` | Interactive provider configuration wizard |\n\n### Smart Agents\n\nMeiGen uses specialized sub-agents for efficient parallel execution:\n\n| Agent | Purpose |\n|-------|---------|\n| `image-generator` | Executes `generate_image` in isolated context — enables true parallel generation |\n| `prompt-crafter` | Writes multiple distinct prompts for batch generation (runs on Haiku for cost efficiency) |\n| `gallery-researcher` | Deep gallery exploration without cluttering the main conversation (runs on Haiku) |\n\n### Output Styles\n\nSwitch creative modes with `\u002Foutput-style`:\n\n- **Creative Director** — Art direction mode with visual storytelling, mood boards, and design thinking\n- **Minimal** — Just images and file paths, no commentary. Ideal for batch workflows\n\n### Automation Hooks\n\n- **Config Check** — Validates provider configuration on session start, guides setup if missing\n- **Auto-Open** — Generated images automatically open in Preview (macOS)\n\n---\n\n\u003Ch2 id=\"providers\">Providers\u003C\u002Fh2>\n\nMeiGen MCP supports three image generation backends. Configure one or multiple — the system auto-selects the best available.\n\n### ComfyUI — Local & Free\n\nRun generation on your own GPU with full control over models, samplers, and workflow parameters. Import any ComfyUI API-format workflow — MeiGen auto-detects KSampler, CLIPTextEncode, EmptyLatentImage, and LoadImage nodes.\n\n```json\n{\n  \"comfyuiUrl\": \"http:\u002F\u002Flocalhost:8188\",\n  \"comfyuiDefaultWorkflow\": \"txt2img\"\n}\n```\n\n> Perfect for Flux, SDXL, or any model you run locally. Your images never leave your machine.\n\n### MeiGen Cloud\n\nCloud API with multiple models: Nanobanana 2, Seedream 5.0, GPT image 1.5, and more. No GPU required.\n\n**Get your API token:**\n1. Sign in at [meigen.ai](https:\u002F\u002Fwww.meigen.ai)\n2. Click your avatar → **Settings** → **API Keys**\n3. Create a new key (starts with `meigen_sk_`)\n\n```json\n{ \"meigenApiToken\": \"meigen_sk_...\" }\n```\n\n### Bring Your Own API (OpenAI-Compatible)\n\nConnect **any** image generation API that follows the OpenAI format — Together AI, Fireworks AI, DeepInfra, SiliconFlow, or your own endpoint. Just provide your key, base URL, and model name:\n\n```json\n{\n  \"openaiApiKey\": \"sk-...\",\n  \"openaiBaseUrl\": \"https:\u002F\u002Fapi.together.xyz\u002Fv1\",\n  \"openaiModel\": \"black-forest-labs\u002FFLUX.1-schnell\"\n}\n```\n\n> All three providers support **reference images**. MeiGen and OpenAI-compatible APIs accept URLs directly; ComfyUI accepts both URLs and local file paths, injecting them into LoadImage nodes in your workflow.\n\n---\n\n## Configuration\n\n### Interactive Setup (Recommended)\n\n```\n\u002Fmeigen:setup\n```\n\nThe wizard walks you through provider selection, API key entry, and ComfyUI workflow import. You can also paste a `curl` command from your API provider's docs — it auto-extracts the key, URL, and model.\n\n### Config File\n\nConfiguration is stored at `~\u002F.config\u002Fmeigen\u002Fconfig.json`. ComfyUI workflows are stored at `~\u002F.config\u002Fmeigen\u002Fworkflows\u002F`.\n\n### Environment Variables\n\nEnvironment variables take priority over the config file.\n\n| Variable | Description |\n|----------|-------------|\n| `MEIGEN_API_TOKEN` | MeiGen platform token |\n| `OPENAI_API_KEY` | Your API key (any OpenAI-compatible provider) |\n| `OPENAI_BASE_URL` | API base URL — change this to use Together AI, Fireworks AI, etc. |\n| `OPENAI_MODEL` | Model name at your provider (e.g., `gpt-image-1.5`, `flux-schnell`) |\n| `COMFYUI_URL` | ComfyUI server URL (default: `http:\u002F\u002Flocalhost:8188`) |\n\n---\n\n## Privacy\n\nMeiGen MCP respects your privacy. Here's what happens with your data:\n\n- **ComfyUI (local)** — All processing stays on your machine. No data is sent externally.\n- **MeiGen Cloud** — Prompts and reference images are sent to `api.meigen.ai` for generation. Generated images are stored temporarily on Cloudflare R2. See [meigen.ai\u002Fprivacy](https:\u002F\u002Fwww.meigen.ai\u002Fprivacy).\n- **OpenAI-compatible** — Prompts and reference images are sent to the configured API endpoint. See your provider's privacy policy.\n- **Reference image upload** — Images are compressed locally (max 2MB) and uploaded to Cloudflare R2 via `gen.meigen.ai`. Uploaded images expire automatically after **24 hours**. No authentication required. ComfyUI users can skip uploading entirely by passing local file paths directly.\n- **Gallery search & prompt enhancement** — Run locally against bundled data. No external API calls.\n\nNo telemetry, analytics, or tracking of any kind.\n\n### Custom Storage Backend\n\nIf you prefer to use your own S3\u002FR2 bucket for reference image uploads, set the `UPLOAD_GATEWAY_URL` environment variable or `uploadGatewayUrl` in `~\u002F.config\u002Fmeigen\u002Fconfig.json` to point to your own presign endpoint. The endpoint must implement:\n\n```\nPOST \u002Fupload\u002Fpresign\nContent-Type: application\u002Fjson\n\nRequest:  { \"filename\": \"photo.jpg\", \"contentType\": \"image\u002Fjpeg\", \"size\": 123456 }\nResponse: { \"success\": true, \"presignedUrl\": \"https:\u002F\u002F...\", \"publicUrl\": \"https:\u002F\u002F...\" }\n```\n\nThe `presignedUrl` is used for a `PUT` upload, and `publicUrl` is the publicly accessible URL returned to the user.\n\n---\n\n## License\n\n[MIT](LICENSE) — free for personal and commercial use.\n","\u003Ch1 align=\"center\">\n  MeiGen AI Design MCP \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fpunkpeye\u002Fawesome-mcp-servers\">\u003Cimg src=\"https:\u002F\u002Fawesome.re\u002Fmentioned-badge.svg\" alt=\"收录于 Awesome MCP Servers\">\u003C\u002Fa> \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwshobson\u002Fagents\u002Ftree\u002Fmain\u002Fplugins\u002Fmeigen-ai-design\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fwshobson%2Fagents-精选-blue?style=flat&logo=github\" alt=\"精选于 wshobson\u002Fagents\">\u003C\u002Fa>\n\u003C\u002Fh1>\n\n\u003Cp align=\"center\">\n  \u003Cstrong>将 Claude Code \u002F OpenClaw 转变为媲美 Lovart 的个人设计助手\u003C\u002Fstrong>\u003Cbr>\u003Csub>本地 ComfyUI 生成，API 集成，1,300+ 专业提示词库，多方向并行输出\u003C\u002Fsub>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002Fmeigen\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fnpm\u002Fv\u002Fmeigen?style=flat-square&color=blue\" alt=\"npm 版本\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002Fmeigen\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fnpm\u002Fdm\u002Fmeigen?style=flat-square&color=green\" alt=\"npm 下载量\">\u003C\u002Fa>\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FType-MCP_服务器-blue?style=flat-square\" alt=\"MCP 服务器\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLocal-ComfyUI-green?style=flat-square\" alt=\"支持 ComfyUI\">\n  \u003Ca href=\"LICENSE\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-lightgrey?style=flat-square\" alt=\"MIT\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FuX6rnersUx\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDiscord-加入-5865F2?style=flat-square&logo=discord&logoColor=white\" alt=\"Discord\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"#quick-start\">快速开始\u003C\u002Fa> •\n  \u003Ca href=\"#see-it-in-action\">演示\u003C\u002Fa> •\n  \u003Ca href=\"#features\">功能\u003C\u002Fa> •\n  \u003Ca href=\"#providers\">提供商\u003C\u002Fa> •\n  \u003Ca href=\"#slash-commands\">命令\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Cstrong>英文\u003C\u002Fstrong> | \u003Ca href=\"README.zh-CN.md\">中文\u003C\u002Fa>\n\u003C\u002Fp>\n\n---\n\n## 这是什么？\n\n这是一个开源的 MCP（Model Context Protocol）服务器（通过插件市场安装），通过 8 个工具和精心设计的技能赋予大语言模型 (LLMs) 创意和审美能力，使其能够处理复杂的设计任务。它教导 LLMs 如何有效使用各种图像生成模型，通过参考图像和多方向并行工作流交付专业结果。\n\n- 与本地 ComfyUI 配合使用——无需外部 API 依赖；也可轻松集成任何自定义 API\n- 内置来自 [nanobanana-trending-prompts](https:\u002F\u002Fgithub.com\u002Fjau123\u002Fnanobanana-trending-prompts) 的 1,500+ 精选提示词模板，以及经过微调的提示词工程技术，可将需求转化为具体的图像生成任务\n- 并行批量生成和子代理执行，以保持主上下文窗口整洁\n\n---\n\n## 演示效果\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fyoutu.be\u002FJQ3DZ1DXqvs\">\n    \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_b53e554df66c.jpg\" alt=\"观看演示\" width=\"400\">\n  \u003C\u002Fa>\n  \u003Cbr>\n  \u003Cb>\u003Ca href=\"https:\u002F\u002Fyoutu.be\u002FJQ3DZ1DXqvs\">▶ 在 YouTube 上观看演示\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fp>\n\n### 产品照片 —— 4 个方向并行\n\n> *\"为这款香水创建 4 张产品展示图，其中一张应包含模特。\"*\n\n**流程** —— AI 上传参考图，构思 4 个不同的提示词，然后并行生成所有 4 张：\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_8982805ea0f6.jpg\" alt=\"并行生成流程\" width=\"700\">\n\u003C\u002Fp>\n\n**结果** —— 不到 2 分钟交付 4 种创意方向：\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_b3ec99310670.jpg\" alt=\"生成结果\" width=\"700\">\n\u003C\u002Fp>\n\n**生成的图片：**\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_6d440009d1b5.jpg\" alt=\"奢华静物\" width=\"24%\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_4a8eb084a9a9.jpg\" alt=\"模特活动\" width=\"24%\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_11ccb9c61940.jpg\" alt=\"自然植物\" width=\"24%\">\n  \u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_readme_6961434888ec.jpg\" alt=\"极简编辑风格\" width=\"24%\">\n\u003C\u002Fp>\n\n---\n\n## 快速开始\n\n### Claude Code 插件（推荐）\n\n```bash\n# 添加插件市场\n\u002Fplugin marketplace add jau123\u002FMeiGen-AI-Design-MCP\n\n# 安装\n\u002Fplugin install meigen@meigen-marketplace\n```\n\n安装后**重启 Claude Code**（关闭并重新打开，或打开新的终端标签页）。\n\n**替代市场** —— 也可通过 [wshobson\u002Fagents](https:\u002F\u002Fgithub.com\u002Fwshobson\u002Fagents)（30k+ stars）获取：\n\n```bash\n\u002Fplugin marketplace add wshobson\u002Fagents\n\u002Fplugin install meigen-ai-design@claude-code-workflows\n```\n\n> 此市场不捆绑 MCP 服务器配置。安装后，将其添加到项目的 `.mcp.json` 中：\n> ```json\n> { \"mcpServers\": { \"meigen\": { \"command\": \"npx\", \"args\": [\"-y\", \"meigen@1.2.6\"] } } }\n> ```\n\n#### 首次设置\n\n免费功能在重启后立即生效——尝试：\n\n> “搜索一些创意灵感”\n\n要解锁图像生成功能，请运行设置向导：\n\n```\n\u002Fmeigen:setup\n```\n\n向导将引导您完成以下步骤：\n1. **选择提供商** —— 本地 ComfyUI、MeiGen Cloud 或任何兼容 OpenAI 的 API（自带密钥和端点）\n2. **输入凭据** —— ComfyUI URL、API Token 或密钥\n3. **完成** —— 再次重启 Claude Code，然后开始生成\n\n### Cursor \u002F VS Code \u002F Windsurf \u002F Roo Code\n\n一条命令即可为任何支持的 AI 编码工具设置 MeiGen：\n\n```bash\nnpx meigen init cursor      # Cursor\nnpx meigen init vscode      # VS Code \u002F GitHub Copilot\nnpx meigen init windsurf    # Windsurf\nnpx meigen init roo         # Roo Code\nnpx meigen init claude      # Claude Code（项目级别）\n```\n\n这将为您的工具写入格式和路径正确的 MCP 配置文件。如果配置文件已存在，MeiGen 将被合并而不会覆盖您的其他服务器。\n\n### OpenClaw\n\n从 [ClawHub](https:\u002F\u002Fclawhub.ai\u002Fjau123\u002Fcreative-toolkit) 安装技能：\n\n```bash\nnpx clawhub@latest install creative-toolkit\n```\n\n或者安装完整插件（包含命令、技能和 MCP 服务器）：\n\n```bash\nopenclaw plugins install meigen-ai-design\n```\n\n### 其他兼容 MCP 的主机\n\n添加到您的 MCP 配置中（例如 `.mcp.json`, `claude_desktop_config.json`）：\n\n```json\n{\n  \"mcpServers\": {\n    \"meigen\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"meigen@latest\"],\n      \"env\": {\n        \"MEIGEN_API_TOKEN\": \"meigen_sk_...\"\n      }\n    }\n  }\n}\n```\n\n> 免费功能（灵感搜索、提示词增强、模型列表）无需任何 API 密钥即可使用。\n\n---\n\n\u003Ch2 id=\"features\">功能\u003C\u002Fh2>\n\n### MCP (模型上下文协议) 工具\n\n| 工具 | 免费 | 描述 |\n|------|------|-------------|\n| `search_gallery` | Yes | 搜索 1,500+ 精选热门提示词，附带视觉预览（由 [nanobanana-trending-prompts](https:\u002F\u002Fgithub.com\u002Fjau123\u002Fnanobanana-trending-prompts) 提供支持） |\n| `get_inspiration` | Yes | 获取任意画廊条目的完整提示词、所有图片及元数据 |\n| `enhance_prompt` | Yes | 将简短想法转化为专业的图像提示词 |\n| `list_models` | Yes | 列出所有已配置提供商的可用模型 |\n| `upload_reference_image` | Yes | 压缩并上传本地图片作为参考 |\n| `comfyui_workflow` | Yes | 管理 ComfyUI 工作流模板：列表、查看、导入、修改、删除 |\n| `manage_preferences` | Yes | 记住您偏好的风格、宽高比、模型和收藏提示词 |\n| `generate_image` | Key | 生成图像 — 自动路由到最佳可用提供商 |\n\n### 斜杠命令\n\n| 命令 | 描述 |\n|---------|-------------|\n| `\u002Fmeigen:gen \u003Cprompt>` | 快速生成 — 跳过对话，直接生成图像 |\n| `\u002Fmeigen:find \u003Ckeywords>` | 搜索 1,300+ 精选提示词以获取灵感 |\n| `\u002Fmeigen:models` | 浏览并切换此会话的 AI 模型 |\n| `\u002Fmeigen:ref \u003Cfile>` | 上传本地图片作为参考，返回 URL |\n| `\u002Fmeigen:setup` | 交互式提供商配置向导 |\n\n### 智能代理\n\nMeiGen 使用专用子代理以实现高效并行执行：\n\n| 代理 | 用途 |\n|-------|---------|\n| `image-generator` | 在隔离上下文中执行 `generate_image` — 实现真正的并行生成 |\n| `prompt-crafter` | 编写多个不同的提示词用于批量生成（在 Haiku 上运行以节省成本） |\n| `gallery-researcher` | 深度探索画廊而不干扰主对话（在 Haiku 上运行） |\n\n### 输出样式\n\n使用 `\u002Foutput-style` 切换创意模式：\n\n- **创意总监** — 视觉指导模式，包含视觉叙事、情绪板和设计思维\n- **极简** — 仅显示图像和文件路径，无评论。适合批量工作流\n\n### 自动化钩子\n\n- **配置检查** — 会话开始时验证提供商配置，若缺失则引导设置\n- **自动打开** — 生成的图像自动在预览程序（macOS）中打开\n\n---\n\n\u003Ch2 id=\"providers\">提供商\u003C\u002Fh2>\n\nMeiGen MCP 支持三种图像生成后端。配置一个或多个 — 系统会自动选择最佳可用项。\n\n### ComfyUI — 本地且免费\n\n在自己的 GPU 上运行生成，完全控制模型、采样器和工作流参数。导入任何 ComfyUI API 格式的工作流 — MeiGen 自动检测 KSampler、CLIPTextEncode、EmptyLatentImage 和 LoadImage 节点。\n\n```json\n{\n  \"comfyuiUrl\": \"http:\u002F\u002Flocalhost:8188\",\n  \"comfyuiDefaultWorkflow\": \"txt2img\"\n}\n```\n\n> 非常适合 Flux、SDXL 或您在本地运行的任何模型。您的图像永远不会离开您的机器。\n\n### MeiGen 云\n\n多模型云 API：Nanobanana 2, Seedream 5.0, GPT image 1.5 等。无需 GPU。\n\n**获取您的 API 令牌：**\n1. 在 [meigen.ai](https:\u002F\u002Fwww.meigen.ai) 登录\n2. 点击头像 → **设置** → **API 密钥**\n3. 创建新密钥（以 `meigen_sk_` 开头）\n\n```json\n{ \"meigenApiToken\": \"meigen_sk_...\" }\n```\n\n### 自带 API（兼容 OpenAI）\n\n连接遵循 OpenAI 格式的**任何**图像生成 API — Together AI, Fireworks AI, DeepInfra, SiliconFlow 或您自己的端点。只需提供密钥、基础 URL 和模型名称：\n\n```json\n{\n  \"openaiApiKey\": \"sk-...\",\n  \"openaiBaseUrl\": \"https:\u002F\u002Fapi.together.xyz\u002Fv1\",\n  \"openaiModel\": \"black-forest-labs\u002FFLUX.1-schnell\"\n}\n```\n\n> 所有三个提供商都支持**参考图像**。MeiGen 和兼容 OpenAI 的 API 直接接受 URL；ComfyUI 接受 URL 和本地文件路径，并将它们注入到您工作流中的 LoadImage 节点。\n\n---\n\n## 配置\n\n### 交互式设置（推荐）\n\n```\n\u002Fmeigen:setup\n```\n\n向导将引导您完成提供商选择、API 密钥输入和 ComfyUI 工作流导入。您也可以从 API 提供商文档粘贴 `curl` 命令 — 它会自动提取密钥、URL 和模型。\n\n### 配置文件\n\n配置存储在 `~\u002F.config\u002Fmeigen\u002Fconfig.json`。ComfyUI 工作流存储在 `~\u002F.config\u002Fmeigen\u002Fworkflows\u002F`。\n\n### 环境变量\n\n环境变量优先级高于配置文件。\n\n| 变量 | 描述 |\n|----------|-------------|\n| `MEIGEN_API_TOKEN` | MeiGen 平台令牌 |\n| `OPENAI_API_KEY` | 您的 API 密钥（任何兼容 OpenAI 的提供商） |\n| `OPENAI_BASE_URL` | API 基础 URL — 更改此地址以使用 Together AI, Fireworks AI 等。 |\n| `OPENAI_MODEL` | 提供商处的模型名称（例如 `gpt-image-1.5`, `flux-schnell`） |\n| `COMFYUI_URL` | ComfyUI 服务器 URL（默认：`http:\u002F\u002Flocalhost:8188`） |\n\n---\n\n## 隐私\n\nMeiGen MCP 尊重您的隐私。以下是关于您数据的处理方式：\n\n- **ComfyUI（本地）** — 所有处理均留在您的机器上。不向外部发送数据。\n- **MeiGen 云** — 提示词和参考图像发送至 `api.meigen.ai` 进行生成。生成的图像临时存储在 Cloudflare R2 上。请参阅 [meigen.ai\u002Fprivacy](https:\u002F\u002Fwww.meigen.ai\u002Fprivacy)。\n- **兼容 OpenAI** — 提示词和参考图像发送至配置的 API 端点。请参阅您提供商的隐私政策。\n- **参考图像上传** — 图像在本地压缩（最大 2MB）并通过 `gen.meigen.ai` 上传至 Cloudflare R2。上传的图像在 **24 小时** 后自动过期。无需身份验证。ComfyUI 用户可以直接传递本地文件路径而完全跳过上传。\n- **画廊搜索与提示词增强** — 针对捆绑数据在本地运行。无外部 API 调用。\n\n无任何遥测、分析或跟踪。\n\n### 自定义存储后端\n\n如果您更喜欢使用自己的 S3\u002FR2 桶进行参考图像上传，请设置 `UPLOAD_GATEWAY_URL` 环境变量或在 `~\u002F.config\u002Fmeigen\u002Fconfig.json` 中设置 `uploadGatewayUrl` 以指向您自己的预签名端点。该端点必须实现：\n\n```\nPOST \u002Fupload\u002Fpresign\nContent-Type: application\u002Fjson\n\nRequest:  { \"filename\": \"photo.jpg\", \"contentType\": \"image\u002Fjpeg\", \"size\": 123456 }\nResponse: { \"success\": true, \"presignedUrl\": \"https:\u002F\u002F...\", \"publicUrl\": \"https:\u002F\u002F...\" }\n```\n\n`presignedUrl` 用于 `PUT` 上传，`publicUrl` 是返回给用户的公开可访问 URL。\n\n---\n\n## 许可证\n\n[MIT](LICENSE) — 个人和商业用途免费。","# MeiGen-AI-Design-MCP 快速上手指南\n\n## 环境准备\n- **Node.js**: 确保系统已安装 Node.js 环境（推荐 LTS 版本）。\n- **包管理器**: 需预装 `npm` 以支持 `npx` 命令。\n- **开发工具**: 支持 Claude Code、Cursor、VS Code、Windsurf、Roo Code 或任何兼容 MCP 协议的主机。\n- **网络**: 需能访问 GitHub 及 npm 公共仓库。\n\n## 安装步骤\n\n### 方案一：Claude Code 插件（推荐）\n在终端中执行以下命令添加市场并安装插件：\n```bash\n\u002Fplugin marketplace add jau123\u002FMeiGen-AI-Design-MCP\n\u002Fplugin install meigen@meigen-marketplace\n```\n安装完成后请重启 Claude Code（关闭并重新打开，或新建终端标签页）。\n\n### 方案二：Cursor \u002F VS Code \u002F Windsurf \u002F Roo Code\n使用一键初始化命令自动配置 MCP 连接：\n```bash\nnpx meigen init cursor      # Cursor\nnpx meigen init vscode      # VS Code \u002F GitHub Copilot\nnpx meigen init windsurf    # Windsurf\nnpx meigen init roo         # Roo Code\nnpx meigen init claude      # Claude Code (项目级)\n```\n此命令会自动写入正确的 `.mcp.json` 配置文件，且不会覆盖现有的其他服务器配置。\n\n### 方案三：手动配置 MCP\n若使用其他 MCP 兼容主机，请在配置文件中添加如下内容：\n```json\n{\n  \"mcpServers\": {\n    \"meigen\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"meigen@latest\"],\n      \"env\": {\n        \"MEIGEN_API_TOKEN\": \"meigen_sk_...\"\n      }\n    }\n  }\n}\n```\n\n## 基本使用\n\n### 首次设置\n启动后运行以下命令进入交互式向导，选择图像生成后端（本地 ComfyUI、MeiGen 云端或自有 API）：\n```\n\u002Fmeigen:setup\n```\n按提示输入凭证（如 ComfyUI URL、API Token 等），完成后再次重启 IDE 即可开始生成。\n\n### 核心指令\n在聊天窗口中可直接调用以下指令：\n- **快速生成**: `\u002Fmeigen:gen \u003C提示词>` —— 跳过对话直接生成图片。\n- **灵感搜索**: `\u002Fmeigen:find \u003C关键词>` —— 搜索 1,300+ 精选提示词。\n- **模型管理**: `\u002Fmeigen:models` —— 浏览并切换当前会话的 AI 模型。\n- **参考图上传**: `\u002Fmeigen:ref \u003C文件路径>` —— 上传本地图片作为参考。\n\n### 示例流程\n1. **搜索灵感**: 输入 `\u002Fmeigen:find luxury perfume`。\n2. **生成图片**: 输入 `\u002Fmeigen:gen Create 4 product display images for this perfume, one of which should feature a model.`。\n3. **并行输出**: 系统会自动分发任务，在 2 分钟内提供多个方向的创意结果。\n\n### 注意事项\n- **免费功能**: 灵感搜索、提示词增强、模型列表无需 API Key 即可使用。\n- **图像生成**: 需完成 `\u002Fmeigen:setup` 配置后端服务。\n- **隐私保护**: 使用本地 ComfyUI 时，所有处理均在本地完成；云端服务数据会发送至 `api.meigen.ai`。","某电商团队正在筹备新品发布会，设计师需要在一小时内产出同一款香水瓶的四种不同风格渲染图，用于社交媒体预热。\n\n### 没有 MeiGen-AI-Design-MCP 时\n- 需要手动在多个绘图平台间切换，反复调整提示词才能出图，打断创作心流\n- 串行生成图片耗时过长，等待一张结果出来后再试下一张，效率极低\n- 缺乏专业提示词库，生成的画面光影和质感往往不符合商业标准，需人工后期修图\n- 依赖第三方 API 导致成本不可控且速度受限，无法处理批量并发任务\n\n### 使用 MeiGen-AI-Design-MCP 后\n- 直接在代码编辑器中通过自然语言指令调用，无需离开开发环境即可完成设计\n- 内置 1,400+ 专业提示词库，自动优化需求转化为高质量生成任务，减少试错\n- 支持多方向并行生成，一次请求即可同时产出四张不同风格的成品，大幅缩短周期\n- 基于本地 ComfyUI 运行，隐私安全且无 API 调用费用限制，适合大规模迭代\n\n核心价值：将复杂的图像生成工作流封装进对话界面，让开发者也能像专业设计师一样高效产出视觉资产。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fjau123_MeiGen-AI-Design-MCP_b53e554d.jpg","jau123",null,"https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fjau123_0456ce48.jpg","meigen7982","https:\u002F\u002Fgithub.com\u002Fjau123",[81,85,89,93],{"name":82,"color":83,"percentage":84},"TypeScript","#3178c6",98,{"name":86,"color":87,"percentage":88},"Shell","#89e051",1.5,{"name":90,"color":91,"percentage":92},"Dockerfile","#384d54",0.3,{"name":94,"color":95,"percentage":96},"JavaScript","#f1e05a",0.1,558,68,"2026-04-04T18:04:52","MIT","未说明",{"notes":103,"python":101,"dependencies":104},"本工具为基于 Node.js 的 MCP Server（通过 npx 安装），核心逻辑不依赖 Python。本地图像生成功能需用户自行部署 ComfyUI 后端（需独立 Python 环境，但版本未说明）。配置文件默认保存在 ~\u002F.config\u002Fmeigen\u002Fconfig.json。支持本地 ComfyUI、MeiGen Cloud 及 OpenAI 兼容 API 三种后端。参考图片上传前会自动压缩至 2MB 以下，且仅保存 24 小时。",[101],[14],[107,108,109,110,111,112,113,114,115,116],"ai-image-generation","claude","claude-code","comfyui","lovart","mcp","mcp-server","model-context-protocol","openclaw","prompt-engineering","2026-03-27T02:49:30.150509","2026-04-06T07:11:49.978778",[120],{"id":121,"question_zh":122,"answer_zh":123,"source_url":124},3360,"如何在项目中获取演示视频的链接？","演示视频已上传至 GitHub 仓库的附件中。维护者提供了直接访问链接，您可以点击以下地址查看：https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F20437f76-f8c8-45f8-a028-c906a568876b","https:\u002F\u002Fgithub.com\u002Fjau123\u002FMeiGen-AI-Design-MCP\u002Fissues\u002F1",[126,131,136,141,146,151,156,161],{"id":127,"version":128,"summary_zh":129,"released_at":130},102946,"v1.2.4","## What's New\n\n- **Gallery categories renamed** — 3D → Illustration & 3D, Food → Food & Drink, Photograph → Photography, Product → Product & Brand (backward-compatible mapping included)\n- **SKILL.md security compliance** — Improved wording to pass VirusTotal\u002FClawHub security scans\n- **Model pricing table** — Added to SKILL.md for quick reference\n- **ClawHub install link** — Added to OpenClaw section in READMEs\n- **Awesome badges** — Updated to Awesome MCP Servers (82k stars), replaced MCP Hub China with Discord invite","2026-03-06T03:07:12",{"id":132,"version":133,"summary_zh":134,"released_at":135},102947,"v1.2.3","## What's New\n\n- **New models** — Added support for Nanobanana 2 and Seedream 5.0 Lite\n- **Default model** — MeiGen provider now defaults to `nanobanana-2` (5 credits, 4K support) when no model is specified\n- **Updated docs** — Model references updated across README, SKILL.md, and plugin metadata","2026-02-27T05:08:15",{"id":137,"version":138,"summary_zh":139,"released_at":140},102948,"v1.2.2","## What's New\n\n- **Semantic search** — `search_gallery` now uses hybrid vector + keyword search, finding conceptually similar results instead of just keyword matches\n- **Automatic fallback** — gracefully falls back to local keyword search if API is unavailable\n- **Glama listing** — added `glama.json` and `Dockerfile` for Glama directory submission\n- **Version sync** — fixed MCP server version reporting across all files","2026-02-26T09:58:27",{"id":142,"version":143,"summary_zh":144,"released_at":145},102949,"v1.2.1","## What's New\n\n- **Semantic search** — `search_gallery` now uses hybrid vector + keyword search via website API, finding conceptually similar results instead of just keyword matches\n- **Automatic fallback** — gracefully falls back to local keyword search if API is unavailable\n- **Glama listing** — added `glama.json` and `Dockerfile` for Glama directory submission\n\n### Install \u002F Update\n```\nnpx -y meigen@latest\n```","2026-02-26T09:55:57",{"id":147,"version":148,"summary_zh":149,"released_at":150},102950,"v1.2.0","## MeiGen v1.2.0\n\nAI design MCP server with multi-provider image generation.\n\n### Features\n- 7 MCP tools: search_gallery, get_inspiration, enhance_prompt, list_models, upload_reference_image, comfyui_workflow, generate_image\n- 3 provider backends: MeiGen Cloud, OpenAI-compatible APIs, local ComfyUI\n- 1,300+ curated prompt library\n- Style-aware prompt enhancement\n- Reference image support across all providers\n- User preferences system\n- Multi-platform init command\n\n### Install\n```\nnpx -y meigen@latest\n```","2026-02-26T09:42:54",{"id":152,"version":153,"summary_zh":154,"released_at":155},102943,"v1.2.7","## Changes\n\n- **Remove `upload_reference_image` tool** — local file paths in `referenceImages` are now auto-compressed and uploaded inline via `generate_image`. Prevents free storage abuse and resolves ClawHub\u002FVirusTotal security scanner flags.\n- **Remove `\u002Fmeigen:ref` command** — no longer needed.\n- **Tool count**: 8 → 7\n- **Update OpenClaw install** to bundle plugin: `openclaw bundles install clawhub:meigen-ai-design`\n\n### How reference images work now\n\nPass local file paths or URLs directly to `generate_image`'s `referenceImages` parameter:\n- **MeiGen \u002F OpenAI-compatible**: local files auto-compressed (max 2MB, 2048px) and uploaded\n- **ComfyUI**: local files passed directly to the workflow (no upload)\n\nNo separate upload step needed.","2026-04-01T10:31:30",{"id":157,"version":158,"summary_zh":159,"released_at":160},102944,"v1.2.6","## What's Changed\n\n### Midjourney Niji 7 Support Improvements\n- **Anime-only guidance**: Niji 7 is exclusively for anime\u002Fillustration styles\n- **Multi-image support**: Niji 7 returns 4 candidate images; MCP server now displays all 4 URLs\n- **Prompt enhancement**: enhance_prompt recommends style anime for Niji 7\n- **SERVER_INSTRUCTIONS**: Added Niji 7 usage notes\n\n### Documentation & Skill Updates\n- Updated model tables in OpenClaw and Claude Code plugin skills\n- OpenClaw skill version 1.0.16\n- Improved release skill coverage\n\n**Full Changelog**: https:\u002F\u002Fgithub.com\u002Fjau123\u002FMeiGen-AI-Design-MCP\u002Fcompare\u002Fv1.2.5...v1.2.6","2026-03-25T02:42:17",{"id":162,"version":163,"summary_zh":164,"released_at":165},102945,"v1.2.5","## What's New\n\n- **24-hour expiry notice** — `upload_reference_image` now clearly informs users that uploaded URLs expire after 24 hours\n- **Magic bytes validation** — File uploads now verify actual image content (JPEG\u002FPNG\u002FWebP\u002FGIF signatures), preventing fake file extensions\n- **ComfyUI local file support** — ComfyUI users can pass local file paths directly to `referenceImages`, skipping the cloud upload roundtrip entirely\n- **OpenAI-compatible provider naming** — All descriptions now say \"OpenAI-compatible\" instead of \"OpenAI\", reflecting support for any compatible API (Together AI, Fireworks AI, etc.)\n- **Custom storage backend docs** — README now includes presign API specification for users who want to use their own S3\u002FR2 bucket\n- **OpenClaw skill restructured** — SKILL.md split into focused core instructions + `references\u002F` directory for provider details and troubleshooting","2026-03-06T03:07:35"]