Transform workflows with AI-powered reliability. Enterprise-grade workflow orchestration powered by Apache Pulsar, LLM routing, and MCP integration.
Project Nexus v2 is an enterprise-grade workflow orchestration platform that transforms how teams build mission-critical AI workflows. Built on Apache Pulsar for reliable messaging and powered by intelligent LLM routing, it enables you to:
- Route messages intelligently with AI-powered classification
- Process documents with AI vision to extract structured data from invoices, receipts, and handwritten notes
- Ensure zero task loss with individual message acknowledgment
- Scale securely with multi-tenant architecture and OAuth2 integration
- Monitor in real-time with Prometheus metrics and Grafana dashboards
- Intelligent Message Routing: Automatically classifies and routes messages to vision, tool execution, or chat workers
- Modern AI Vision: Layout-aware OCR with table preservation for structured data extraction
- Pulsar Backbone: Individual message acknowledgment ensures no task is lost, even when workers fail
- Multi-Tenancy: OAuth2 integration with isolated Pulsar tenants and namespaces for enterprise-grade security
- MCP Integration: Execute tools via Model Context Protocol (Stripe payments, database operations, custom workflows)
- Real-Time Monitoring: Track worker status, consumer lag, and task completion with Prometheus and Grafana
Project Nexus v2 includes support for the following MCP servers:
- ✅ Exa Search: Live web search using Exa MCP API (HTTP transport)
- ✅ Google Maps Grounding: Location search and mapping services (HTTP transport)
- ✅ Playwright: Browser automation and web scraping (stdio transport with headless mode)
- ✅ GitHub: GitHub repository and code management (stdio transport)
- ✅ LangChain Agent: Multi-step reasoning and tool orchestration via LangChain ReAct agents (HTTP transport)
- ✅ Google Workspace: Comprehensive GSuite integration (Gmail, Drive, Calendar, Docs, Sheets, Slides, Forms, Tasks, Chat) via stdio transport
- ✅ Notion: Notion workspace management, page search, database operations, and knowledge organization (stdio transport)
- ✅ n8n Automation: Visual workflow building, triggers, and integrations with thousands of services (stdio transport)
- ✅ Sequential Thinking: Chain-of-thought reasoning and structured problem-solving for AI agents (stdio transport)
All system servers are pre-configured and can be enabled/disabled through the monitoring dashboard. Custom MCP servers can be added via the server management interface.
Note: Some servers require environment variables for authentication:
- Google Workspace:
GOOGLE_OAUTH_CLIENT_IDandGOOGLE_OAUTH_CLIENT_SECRET - Notion:
NOTION_API_KEY - GitHub: GitHub Personal Access Token (via settings or environment variable)
- Google Maps:
GOOGLE_MAPS_GROUNDING_API_KEY
- Framework: Next.js 16 (React 19)
- Language: TypeScript 5
- Styling: Tailwind CSS v4
- UI Components: Radix UI, shadcn/ui
- Package Manager: pnpm
- Database: Supabase (PostgreSQL)
- Backend: Supabase Edge Functions (Deno)
- Message Queue: Apache Pulsar
- Deployment: Vercel
- Node.js 18+
- pnpm 10.x (or use
corepack enableto manage it automatically) - Supabase account (for database and authentication)
- Apache Pulsar instance (or use managed Pulsar service)
-
Clone the repository:
git clone https://github.com/mcpmessenger/project-nexus-v2.git cd project-nexus-v2-main -
Install dependencies:
pnpm install
-
Set up environment variables:
cp .env.example .env.local
Edit
.env.localwith your configuration values (Supabase URL, API keys, Pulsar connection details, etc.). -
Run database migrations:
# Follow instructions in docs/STORAGE_SETUP.md -
Run the development server:
pnpm dev
-
Open http://localhost:3000 in your browser.
pnpm dev- Start development serverpnpm build- Build for productionpnpm start- Start production serverpnpm lint- Run ESLint
├── app/ # Next.js app directory
│ ├── api/ # API routes
│ ├── monitoring/ # Registry/MCP server management
│ └── workflows/ # Workflow orchestration UI
├── components/ # React components
│ ├── ui/ # Reusable UI components
│ └── workflow-diagram.tsx # Visual workflow architecture
├── lib/ # Utility functions and clients
├── public/ # Static assets
│ └── images/ # Server logos and images
├── supabase/ # Supabase configuration
│ ├── functions/ # Edge functions
│ └── migrations/ # Database migrations
└── docs/ # Architecture and documentation
This project is automatically deployed to Vercel when changes are pushed to the main branch.
- Repository:
mcpmessenger/project-nexus-v2 - Branch:
main - Platform: Vercel
To trigger a manual rebuild, push a commit to main or trigger a redeploy from the Vercel dashboard.
Run the following commands to validate your setup:
pnpm lint # Validate TypeScript and lint rules
pnpm build # Ensure production bundle compilesComprehensive documentation is available in the docs/ directory:
- Architecture: System design and component architecture
- Architecture v2: Updated architecture with improvements
- Implementation Roadmap: Planned features and milestones
- Operational Guide: Running and maintaining the system
- Storage Setup: Database and storage configuration
- Setup Guide: Detailed setup instructions
We welcome contributions! Please see our Contributing Guide for detailed guidelines. Here's a quick overview:
- Fork the repository
- Create a feature branch from
main:git checkout -b feature/your-feature-name
- Make your changes
- Test thoroughly:
pnpm lint # Check for linting errors pnpm build # Ensure everything compiles
- Commit your changes with a clear message:
git commit -m "feat: Add your feature description" - Push to your fork:
git push origin feature/your-feature-name
- Open a pull request on GitHub
- Follow the existing code style and conventions
- Add tests for new features when applicable
- Update documentation if you change functionality
- Write clear, descriptive commit messages
- Ensure your code passes linting and builds successfully
- MCP Server Adapters: Implement additional MCP server integrations
- UI/UX Improvements: Enhance the user interface and experience
- Documentation: Improve docs, add examples, or fix typos
- Testing: Add unit tests, integration tests, or E2E tests
- Performance: Optimize workflows, reduce latency, improve scalability
- Bug Fixes: Report and fix issues
For major changes, please open an issue first to discuss what you'd like to change.
See LICENSE file for details.
- Issues: Open an issue on GitHub
- Discussions: GitHub Discussions
- Documentation: Check the
docs/directory for detailed guides
Built with ❤️ by the Project Nexus team