I'm excited to share an open-source project that bridges the gap between modern AI and MultiValue databases.
u2-mcp is a Model Context Protocol (MCP) server that enables AI assistants like Claude to directly query and interact with Rocket Universe and UniData databases—while preserving native MultiValue semantics.
🔑 Key Features:
• Query files using RetrieVe/UniQuery syntax
• Read/write records with full dynamic array support
• Access dictionaries and call BASIC subroutines
• Built-in safety controls (read-only mode, command blocklist, query limits)
• OAuth authentication for secure cloud deployment
💡 Why This Matters:
Instead of flattening MultiValue data into relational models, u2-mcp preserves attributes, multivalues, and subvalues natively. Your AI assistant understands your data the way your applications do.
Use Cases:
• Natural language queries against your ERP/business data
• AI-assisted report generation
• Database documentation and exploration
• Training new developers on legacy systems
The project uses Rocket's official uopy package and is available now on PyPI and GitHub.
📦 pip install u2-mcp
🔗 GitHub: github.com/bpamiri/u2-mcp
I'd love to hear from others in the MultiValue community. What would you want an AI assistant to help with in your U2 environment?
