azure-content-understanding
MCP server from Azure-Samples/azure-content-understanding-mcp-server
claude mcp add --transport stdio azure-samples-azure-content-understanding-mcp-server dotnet run --project FULL_PATH_TO_PROJECT_FOLDER --no-build \ --env API_KEY="CONTENT_UNDERSTANDING_API_KEY" \ --env ENDPOINT="CONTENT_UNDERSTANDING_ENDPOINT" \ --env API_VERSION="CONTENT_UNDERSTANDING_VERSION" \ --env ALLOWED_DIRECTORIES="SEMICOLON_SEPERATED_LIST_OF_FOLDERS" \ --env STORAGE_CONTAINER_URL="STORAGE_CONTAINER_URL"
How to use
This MCP server exposes Azure Content Understanding capabilities through a standard MCP interface. It provides a ContentUnderstanding toolset that allows clients to create and manage analyzers, submit documents for analysis, and retrieve results. The server handles document upload to Azure Blob Storage, submission to the Content Understanding API, polling for analysis completion, returning structured results, and cleaning up temporary storage. With multiple analyzers supported, you can tailor analysis workflows to content type or data requirements, making it suitable for diverse document analysis scenarios across enterprise content pipelines.
How to install
Prerequisites:
- .NET 9.0 SDK
- An Azure subscription with the Content Understanding service
- An Azure Blob Storage account
Installation steps:
- Install the .NET runtime and SDK if not already installed. For example:
- On Windows: install via official .NET download page
- On macOS/Linux: use your package manager or official installers
- Clone or download the MCP server repository for Azure Content Understanding.
- Build the project locally (if needed) or run directly with the provided configuration. If you have a project path, replace FULL_PATH_TO_PROJECT_FOLDER in the config.
- Ensure you have an Azure Content Understanding endpoint and API key, and a configured storage container URL. Populate the environment variables in the mcp_config section accordingly.
- Run the MCP server using the configured command from the mcp_config. Example command provided in the configuration below.
Additional notes
Tips and common issues:
- Ensure your Azure credentials (endpoint and API key) are valid and have permissions for the Content Understanding service.
- The ALLOWED_DIRECTORIES environment variable should be a semicolon-separated list of folders if you want to restrict analysis scopes.
- If you modify the project path, update FULL_PATH_TO_PROJECT_FOLDER accordingly in the configuration.
- The server expects Azure Blob Storage for temporary uploads; verify storage container access policies and network connectivity.
- Monitor for rate limits on the Azure Content Understanding API and implement proper retry/polling strategies in your integration.
Related MCP Servers
mssql
MSSQL Server MCP implementation written in C#
McpDotNet.Extensions.SemanticKernel
Microsoft SemanticKernel integration for the Model Context Protocol (MCP). Enables seamless use of MCP tools as AI functions.
DatabaseMcpServer
MCP server from ttcc666/DatabaseMcpServer
mcp-dataverse
MCP Server for querying Dataverse using SQL
xperience-community
ASP.NET Core MCP server for Xperience by Kentico projects
console-to-http
Example of converting a stdio MCP server to HTTP using ModelContextProtocol.AspNetCore