loading…
Search for a command to run...
loading…
A production-grade MCP server that turns natural language intent into fully-architected, accessible, production-ready UI code through a 7-step agentic pipeline.
A production-grade MCP server that turns natural language intent into fully-architected, accessible, production-ready UI code through a 7-step agentic pipeline.

CI PyPI Node.js React Python 3.10+ MCP License: MIT Benchmark
VibeServe is the first full-stack, fully-integrated AI development platform. It unifies a React-based AI IDE, a Node.js Orchestrator, and a Python MCP Server into a single cohesive system.
It transforms natural language intent into fully-architected, accessible, production-ready UI code — powered by a 7-step agentic pipeline, a fully aware context engine, and dynamic human-in-the-loop controls.
🏗️ architect → 💻 code → 🔍 review → ✅ verify → 🔄 iterate → 🧪 test → 🚀 deploy
This repository is a monorepo housing the three core pillars of the VibeServe ecosystem:
ide/ (The Front-End Client)The VibeServe IDE (formerly Nexus-Alpha) is a Vite/React application acting as the command center. It features an interactive command palette, real-time agent trajectories, code editors, artifact previews, and full MCP tool visibility. It's fully WCAG AAA accessible and designed for extreme performance.
orchestrator/ (The Control Plane)The VibeServe Orchestrator (formerly CodeNexus) is a high-performance Node.js backend. It coordinates agents, handles WebSockets for real-time streaming, runs deep security audits (AST analysis), and manages the lifecycle of the AI agents and their sandbox environments.
mcp/ (The Python Backend)The original VibeServe MCP Server. This is a production-grade FastMCP server containing 28 specialized tools, SQLite-backed memory, multi-LLM routing with auto-fallback (OpenAI, DeepSeek, OpenRouter, Local Ollama), and direct integration endpoints.
| Feature | Description |
|---|---|
| True Full-Stack AI | Complete visibility from IDE UI down to the python MCP tool execution. |
| Agentic Pipeline | 7 discrete steps (architect to deploy), each chainable or runnable standalone. |
| Auto-Fallback LLMs | Built-in router across 6 providers prevents single point of failure. |
| Trinity Benchmarked | System-wide automated auditing and benchmarking scores ensuring A-grade performance and security. |
| Human-in-the-Loop | You choose when to intervene. The IDE allows seamless manual takeover of any automated workflow. |
Clone the monorepo:
git clone https://github.com/ncsound919/VibeServe
cd VibeServe
Start the MCP Server:
cd mcp
pip install -e ".[dev]"
cp .env.example .env
# Start in dev mode or hook into Claude Desktop
pytest tests/ -v
Start the Orchestrator:
cd ../orchestrator
pnpm install
pnpm run build
pnpm run dev
Start the IDE:
cd ../ide
npm install
npm run dev
The system will now be fully interconnected.
VibeServe IDE (React / Vite)
↓ WebSocket / REST
VibeServe Orchestrator (Node.js / Express)
↓ stdio MCP Protocol
VibeServe MCP Server (Python / FastMCP)
├── 28 Tools · 5 Resources · 6 Prompts
├── V5 Agentic Pipeline
├── LLMRouter (Auto-fallback chain)
├── MemoryStore (aiosqlite)
└── SchemaValidator (WCAG AAA enforcement)
The entire VibeServe ecosystem undergoes a rigorous, automated audit (The Trinity Benchmark) across Performance, Code Quality, Security, and Features.
Current System Score: 91.9/100 (Grade A)
Run the benchmark yourself:
cd mcp
python benchmark_system.py
VibeServe is free and open source. If it saves you time:
💚 CashApp: $helptools
Every dollar helps keep the tools free.
MIT License. See LICENSE for details.
Добавь это в claude_desktop_config.json и перезапусти Claude Desktop.
{
"mcpServers": {
"vibeserve": {
"command": "npx",
"args": []
}
}
}PRs, issues, code search, CI status
автор: GitHubDatabase, auth and storage
автор: SupabaseReference / test server with prompts, resources, and tools.
Secure file operations with configurable access controls.
Не уверен что выбрать?
Найди свой стек за 60 секунд
Автор?
Embed-бейдж для README
Похожее
Все в категории development