Skip to content

Latest commit

 

History

History
395 lines (277 loc) · 15.6 KB

File metadata and controls

395 lines (277 loc) · 15.6 KB

🚀 MCP-X — Enterprise AI Platform

All-in-One AI Creative Workflow · Chat / Video / Image / Code / Agent / MCP

License GitHub stars GitHub forks React TypeScript Vite TailwindCSS

🌐 Live Demo · 📖 API Docs · 💬 Join Community · 🐛 Report Issue

🇨🇳 中文版


🔥 Limited Offer: Join the Group for Free 100 RMB Image & Video Tokens

🎉 Now Live: GPT Image 2 and Stable Diffusion 2.0 are fully available on MCP-X! Top-tier international image generation models, all in one place.

💰 New User Exclusive: Register now and get 100 RMB in free credits — try GPT Image 2, SD 2.0, Jimeng, Kling and more. Go wild with the world's best image models!

🎬 Video Generation Tokens  |  🖼️ Image Generation Tokens  |  ⚡ Instant Delivery

WeChat Group

👆 Scan to Join & Claim Now


📸 Product Preview

Unified Portal
Unified Portal
Featured Content
Featured Content
Image Studio
Image Studio
Video Studio
Video Studio

🎬 AIGC Video Showcase — Made with MCP-X Video Studio

The following videos were fully produced by MCP-X Video Studio — from script writing, storyboard design, keyframe generation to final output, entirely AI-driven.

🗡️ Legend of Sword and Fairy

Legend of Sword and Fairy - AI Generated Video
▶️ Click to Watch

Chinese Fantasy Theme · AI Script Analysis · Multi-scene Storyboard · Character Design

⚡ Thunder Strike No.1

Thunder Strike No.1 - AI Generated Video
▶️ Click to Watch

Sci-Fi Action Theme · Full AI Pipeline · Video Compositing & Export

💡 Click the cover image to watch on GitHub. You can also visit the MCP-X Platform for the full video creation experience.


✨ Why MCP-X?

🎬
End-to-End Video Production
From script to final cut, AI-assisted
Supports Seedance 2.0 / Veo3 / Kling etc.
🖼️
Professional Image Processing
Text-to-Image / Image-to-Image / Inpainting
Multiple models for all creative needs
💻
AI Code Generation
Conversational frontend dev, live preview
Supports React / Vue / HTML
🤖
Smart Chat
GPT / DeepSeek / Kimi etc.
MCP Tool Calls + Knowledge Base RAG
🔧
MCP Marketplace
1000+ MCP Tools
One-click setup, plug and play
🧠
Agent Marketplace
500+ Vertical Agents
Ready-to-use industry solutions
📚
Enterprise Knowledge Base
Private document retrieval
RAG-enhanced, precise Q&A
🌍
i18n & Private Deployment
Bilingual (CN/EN) interface
Supports on-premise deployment

🎨 Latest Updates

v0.1.8 — Homepage Load Speed · Canvas Cache Fix · Privacy Compliance
  • 🏠 Homepage load experience: Significantly reduced black screen on refresh
    • Homepage eager-loaded (no route lazy-load wait); inline HTML skeleton + hero.webp preload
    • Showcase list sessionStorage cache for instant display on refresh
    • Lazy-loaded cards with skeleton placeholders; hero.mp4 only loads in video mode
  • 🖼️ Image Editor · canvas cache fix: Fixed images/videos failing to load after refresh with many elements or after video generation
    • Filter temporary signed URLs (volces / DashScope); prefer permanent OSS links and base64
    • Merge missing canvas elements from chat history on refresh; graceful localStorage quota fallback
    • Image API timeout extended to 30 minutes; OSS URLs no longer re-fetched as base64
    • Media load concurrency control and serial video restore to prevent resource starvation
  • 📜 Privacy policy compliance: Added OAID, ANDROID ID, location, and AdScope SDK disclosures for Huawei AppGallery review
v0.1.7 — Storyboard Mode · Image 2 · One-Click Web Generation
  • 🎬 Video Studio · Storyboard Mode: Professional storyboard workflow
    • GPT Image 2 generates multi-panel storyboards from scene/character references and shot scripts
    • Seedance 2.0 reference-based video generation preserves character and scene consistency
    • Concurrent multi-shot generation with per-task status and retry on failure
    • Best-in-platform video pipeline: controllable storyboards → reference-guided video
  • 🖼️ Image Studio · GPT Image 2: Full integration of Image 2 — the strongest image model on the platform
  • 💻 App Builder · One-Click Web: Improved stability and preview for AI-generated web pages
v0.1.6 — HTML to PPT Export & App Builder Enhancements
  • 📊 HTML to PPT Export: App Builder now supports one-click export of AI-generated HTML slides to .pptx format
    • Background and text separation: screenshots serve as background layers, text is overlaid as independent editable text boxes in PowerPoint / WPS
    • Smart font mapping: auto-detects Chinese/English content and maps to system fonts (Microsoft YaHei, SimSun, Calibri, etc.) for consistent cross-platform rendering
    • Auto slide detection: supports section[data-title], .slide, [data-slide] and other common HTML presentation frameworks
    • Full-screen loading overlay during export to prevent accidental interactions
  • 🔧 Export Stability Fix: Resolved duplicate page issue caused by stale script listeners from previous export sessions — fixed via version guard + listener reference management
  • 💡 App Builder UX: Export button only enabled after preview page is fully loaded; cross-origin previews show a clear prompt
v0.1.5 — Video Pipeline & Cache Experience Optimization
  • 🎬 Seedance 2.0 Support: New video generation model with significantly improved quality and motion
  • 🖼️ Canvas Video Thumbnail Fix: Fixed intermittent missing video covers after page refresh
  • 🔒 Global 401 Interceptor: Unified session timeout handling across all API services with auto-redirect to login
  • 📋 Paste @Tag Auto-linking: Pasting text with @ImageN / @VideoN tags auto-matches canvas media elements
v0.1.3 — App Builder & Preview
  • 🍌 Nano Banana Pro 4K: New high-resolution image generation model
  • 🔒 HTTPS Preview Proxy: SSE-based preview path parsing, same-protocol preview to avoid mixed content
  • 🖼 Cross-origin Preview Editing: fetch + srcdoc injection for editor scripts, compatible with cross-origin dev preview
  • 📋 New App Page: Displays "My Apps" list with navigation to build page
  • ✏️ Builder UX: 8000-char input limit, source code tab supports inline editing
v0.1.2 — Docs & Login Page Upgrade
  • 📚 Comprehensive documentation upgrade, adapted for OpenClaw integration
  • 🔐 Login page redesign: glassmorphism + dynamic glow effects
  • 🎬 Image-to-video enhancement: selected image auto-used as start frame
v0.1.1 — Reference-based Video Generation
  • 🎬 Support @ symbol to reference canvas images for video generation
  • ✨ Full UI optimization

🌟 Core Features

🎬 1. AI Video Studio

Supports: Qwen / Jimeng / Kling / Hailuo / Seedance 2.0 / Veo3 / Runway etc.

End-to-end AI video production from script to final cut:

Feature Description
Storyboard Mode GPT Image 2 storyboards → Seedance 2.0 reference video — recommended for highest quality
Script Analysis AI auto-analyzes scripts, extracts characters, scenes, story segments
Storyboard Generation Smart storyboard with camera movement, shot types, keyframes
Characters & Scenes AI-generated portraits and scene art as references for storyboard & video
Video Generation Text-to-video / Image-to-video / Keyframe interpolation / Reference mode
One-click Export Browser-based video segment merging and export
Asset Library IndexedDB local storage, reusable character/scene/video assets

💬 2. AI Chat System

Supports: GPT / Gemini / DeepSeek / Kimi / Qwen etc.

  • Multi-model switching · SSE streaming · MCP tool calls
  • Web search integration · Agent workflow with human-in-the-loop · File upload chat
  • Enterprise knowledge base with RAG-enhanced retrieval

💻 3. Frontend App Builder

Similar to Bolt / Loveable AI frontend building experience

  • One-click web generation: describe your page in natural language, get a full preview-ready site instantly
  • Conversational development with multi-turn layout and style iteration
  • Multi-framework: HTML / React / Vue / Static sites
  • Live preview · Visual editing · One-click deploy · Code download · HTML to PPT export

🖼️ 4. AI Image Editor

Supports: Qwen / Seed / Nano Banana / GPT Image 2 / Nano Banana Pro 4K etc.

  • GPT Image 2: strongest image model on the platform for text-to-image and image-to-image
  • Text-to-image · Image-to-image · Mask inpainting
  • Image-to-video · @ symbol reference-based video generation
  • Multi-model selection · Multiple size options

🔧 5. MCP Marketplace & Agent Marketplace

  • 1000+ MCP tools, browse by category, search, one-click setup
  • 500+ Agent bots, featured/latest/categorized, ready to use

🏢 Enterprise Features

Feature Description
🔐 Security Token authentication · Multi-tenant architecture · Data isolation
Performance SSE streaming · IndexedDB local cache · Async task processing
🔌 Open Integration MCP protocol · Unified multi-model API · RESTful API
📊 Management Knowledge base · Workflow orchestration · Human approval · Usage billing
🌐 Internationalization Bilingual (CN/EN) · On-premise deployment · Data compliance

🚀 Quick Start

Requirements

  • Node.js 18+
  • npm or yarn

Install & Run

# Clone the project
git clone https://github.com/TimeCyber/MCP-X-web.git
cd MCP-X-web

# Install dependencies
npm install

# Development mode
npm run dev

# Production build
npm run build

Configuration

Modify the API URL in src/config/index.ts:

const apiBaseUrl = import.meta.env.VITE_API_BASE_URL ||
  (import.meta.env.PROD ? 'https://www.mcp-x.com/prod-api' : 'https://www.mcp-x.com/prod-api');

🛠 Tech Stack

Category Technology
Frontend Framework React 18 + TypeScript
Build Tool Vite 5
UI Tailwind CSS + Ant Design
State Management Zustand + React Context
Routing React Router v6
HTTP Client Axios
Markdown react-markdown + remark-gfm
Animation Framer Motion
Icons Lucide React
Local Storage IndexedDB
Video Processing FFmpeg.wasm
File Compression JSZip

📁 Project Structure

src/
├── components/      # Components
│   ├── app/         # App Builder
│   ├── chat/        # Chat
│   ├── image-editor/# Image Editor
│   ├── layout/      # Layout
│   ├── ui/          # Common UI
│   ├── video-gen/   # Video Generation
│   └── workflow/    # Workflow
├── config/          # Configuration
├── contexts/        # React Context
├── hooks/           # Custom Hooks
├── i18n/            # Internationalization
├── pages/           # Pages
├── services/        # API Services
├── types/           # TypeScript Types
└── utils/           # Utilities

📖 Documentation


🚀 Backend Open Source Plan

⭐ 1000 Stars = Full Backend Open Source

Item Details
Current Status Backend code being organized, open source coming soon
Conditions GitHub reaches 1000 ⭐ + 300 real developers in WeChat group
Backend Stack Spring Boot 3.x · MySQL 8.0 · Redis · Docker & K8s
AI Integration Qwen · Jimeng · Seedance · Veo3 etc.

Support us: ⭐ Star this project · 🔄 Share to tech communities · 💬 Join discussions


📄 License

This project is licensed under Apache License 2.0 with additional commercial terms:

Use Case Free?
✅ Individual users Free
✅ Educational institutions Free
✅ Non-profit organizations Free
✅ Companies with fewer than 20 employees Free
⚠️ Companies with 20+ employees Commercial license required

📧 Commercial licensing: ganyizhi@timecyber.com.cn  |  See LICENSE


🤝 Contact Us

WeChat Group

Scan to join the MCP-X Tech Community


📜 Intellectual Property Notice

⚠️ This project has registered software copyright. MCP-X is a registered trademark. Licensed under Apache License 2.0 with all trademark and copyright rights reserved. Unauthorized commercial use is prohibited. Respect intellectual property and help maintain a healthy open source ecosystem.