OmniXy

OmniXy: Universal Model Context Protocol (MCP) Client

OmniXy is a versatile Model Context Protocol (MCP) client designed for seamless integration with Large Language Models (LLMs). It supports structured reasoning through chain-of-thought and manages prompts, responses, and contexts according to the MCP specification.

Key features include vendor-agnostic design, standards-based protocol, extensibility via a driver pattern, reasoning capabilities, security, scalability, and type safety using Pydantic. It supports various LLM providers, with model names structured as provider-model-name:version.

Getting started involves installation via uv pip install -r requirements.txt, configuration of LLM provider credentials, and usage examples for basic completion, chain-of-thought reasoning, multiple provider integration, custom provider configuration, and error handling. The project structure includes core components, drivers for different LLMs, orchestration tools, and a plugin system. Comprehensive testing with unit and integration tests ensures code quality and coverage.