Awesome MCP › Agent Orchestration Platforms

Abiorh001/mcp_omni_connect

⭐ 245 Python added to this list on 2025-05-14 repository created 2025-03-19

MCPOmni Connect is a universal command-line interface (CLI) client designed to connect to multiple Model Context Protocol (MCP) servers, integrating various AI models and transport protocols to enable advanced task execution, tool orchestration, and resource management within the MCP ecosystem. It functions as an intelligent AI agent system with multiple operation modes including ReAct Agent Mode for autonomous reasoning and decision-making, Orchestrator Mode for strategic multi-step task planning and execution across multiple MCP servers, and Interactive Chat Mode for human-in-the-loop task execution with approval workflows and educational guidance. The project supports multi-protocol connectivity such as stdio transport, Server-Sent Events (SSE), streamable HTTP, Docker container integration, and NPX package execution, with extensible transport layers for future protocols. It also includes robust authentication support with OAuth 2.0, bearer tokens, and custom headers, ensuring secure credential management. MCPOmni Connect offers seamless switching between different agentic operation modes with persistent state management. A key feature is its AI-powered intelligence through unified LLM integration with LiteLLM, supporting over 100 models from providers like OpenAI, Anthropic, Google, Groq, DeepSeek, Azure OpenAI, OpenRouter, and Ollama. It simplifies configuration, manages dynamic system prompts, and enables intelligent context management and automatic tool selection and chaining. The system also supports models without native function calling by dynamically executing functions based on user requests. Security and privacy are prioritized with explicit user control over tool executions, strict data access controls, server-specific data isolation, encrypted communication, and a privacy-first approach that minimizes data collection and prevents unauthorized data exposure. Memory management is handled via Redis-powered persistence for long-term conversation storage and session persistence, along with chat history file storage for saving and loading conversations. Intelligent context management ensures relevant information retrieval and memory-aware responses. Advanced prompt management features include dynamic prompt discovery, flexible argument parsing, cross-server prompt coordination, intelligent validation, and real-time prompt responses. Tool orchestration capabilities allow dynamic tool discovery, cross-server coordination, and intelligent tool selection. Resource management supports universal resource access, cross-server discovery, and smart content summarization. Server management includes multiple simultaneous connections, health monitoring, graceful connection handling, and runtime configuration updates. Overall, MCPOmni Connect is a comprehensive, intelligent, and secure universal gateway for interacting with the MCP ecosystem, enabling sophisticated AI-driven task execution and orchestration across diverse servers and models.

https://github.com/Abiorh001/mcp_omni_connect

ai-agent-systemai-integrationai-modelsanthropicautonomous-reasoningazure-openaibearer-tokenchat-historycli-clientdeepseekdeepseek-modelsdocker-integrationencrypted-communicationgemini-modelsgooglegroqgroq-modelsinteractive-chat-modelitellmmcpmemory-managementmodel-context-protocolmulti-protocol-supportnpx-packageoauth-2.0ollamaopenaiopenai-modelsopenrouteropenrouter-modelsorchestrator-agent-modeorchestrator-modeprivacyprompt-managementreact-agent-modereact-agentic-moderedisredis-persistenceresource-managementsecure-credential-managementsecurityserver-managementserver-sent-eventsssestdio-transportstreamable-httptask-executiontool-managementtool-orchestrationunified-llm-integrationworkflow-management