Overview
Azure Model Context Protocol (MCP) Hub is a central resource for building, running, and reusing MCP servers on Azure to power AI agents. It guides you to run your own MCP server using Azure Functions across C#, Python, or TypeScript, and to build clients, servers, or tools with official MCP SDKs in C#, Python, TypeScript, and Java. It highlights AI SDK integrations to connect agents to MCP servers via interfaces like OpenAI Agents SDK, Semantic Kernel, LangChain.js, Java Spring AI, and Azure AI Agents. For convenience, the hub also showcases plug-and-play MCP servers that expose real APIs via MCP; these servers cover data APIs (Redis, PostgreSQL, MySQL, MongoDB, Azure Data Explorer) and development/infra APIs (Azure CLI, Kubernetes, GitHub, Azure DevOps in Python and Node). The repository stresses that these servers are open source and community-maintained, with Microsoft not reviewing or endorsing them. Learn more via the MCP Spec and linked repos, explore sample agents, tools, and servers, and contribute via PR.
Features
Run MCP servers on Azure Functions
Host your MCP server using Azure Functions with language options C#, Python, or TypeScript.
Official MCP SDKs for building clients, servers, or tools
Leverage official MCP SDKs in C#, Python, TypeScript, and Java to build clients, servers, or tools.
AI SDK integrations
Integrate MCP servers with AI frameworks and tools (OpenAI Agents SDK, Semantic Kernel, LangChain.js, Spring AI, Azure AI Agents).
Plug-and-play MCP servers
Access community-maintained MCP servers exposing real APIs for agents to call.
Data API MCP servers
MCP servers for data stores such as Redis, PostgreSQL, MySQL, MongoDB, and Azure Data Explorer.
Dev & Infra MCP servers
MCP servers for developer tools and infra APIs like Azure CLI, Kubernetes, GitHub, and Azure DevOps (Python/Node.js).
Learn and contribute
Consult the MCP Spec, explore sample agents/tools/servers, and contribute via PR.
Who Is This For?
- Azure developers:Build, run, or reuse MCP servers on Azure to power AI agents.
- Multi-language developers:Build clients, servers, or tools across languages using official MCP SDKs.
- AI engineers:Connect agents to MCP servers with SDKs and integrations efficiently.
