scicomp-mcp-common 0.1.5

Shared infrastructure for MCP servers: GPU management, async tasks, config, serialization

scicomp-mcp-common

Common utilities and base classes for the Math-Physics-ML MCP system.

Overview

This package provides shared functionality used across all MCP servers in the Math-Physics-ML system, including:

• MCP server base classes and utilities
• Common error handling patterns
• Shared type definitions and protocols
• Utility functions for tool registration and management

Installation

pip install scicomp-mcp-common

Usage

This package is primarily used as a dependency by other MCP servers in the system. Direct usage is typically not needed for end users.

from mcp_common import MCPServer, Tool

# Example: Used by other MCP servers
server = MCPServer()

Part of Math-Physics-ML MCP System

This package is part of a larger system that includes:

• scicomp-compute-core - GPU-accelerated computation core
• scicomp-math-mcp - Symbolic math and numerical computing
• scicomp-quantum-mcp - Quantum mechanics simulations
• scicomp-molecular-mcp - Molecular dynamics
• scicomp-neural-mcp - Neural network training

See the full documentation for more details.