MCP Server for ReportServer Java Application

What is MCP Server for ReportServer Java Application?

An MCP (Model Context Protocol) server implementation that connects to a Java ReportServer application, providing a chat agent interface for report management and generation. It uses .NET 9.0 with Aspire orchestration, Microsoft Extensions AI SDK, Ollama for AI models, and Qdrant for vector search—intended for developers building AI-powered reporting tools.

How to use MCP Server for ReportServer Java Application?

Start the development environment by navigating to the RSChatApp.AppHost directory and running dotnet run. This launches all containerized services (MCP server, Blazor chat web app, Ollama, Qdrant) orchestrated via .NET Aspire. Access the Aspire dashboard (default http://localhost:15986) and the chat web interface (default http://localhost:5123) to interact. For a quick functional test, run the provided test-mcp-server.sh script.

Key features of MCP Server for ReportServer Java Application

• Uses official Microsoft Extensions AI SDK
• Full .NET 9.0 integration with Aspire orchestration
• Direct ReportServer RPC integration
• Comprehensive logging and error handling

Use cases of MCP Server for ReportServer Java Application

• Query available reports through an AI-powered chat interface
• Generate and export reports by conversing with the AI assistant
• Manage report tree and properties via natural language requests
• Integrate report functionality into VS Code via independent MCP server deployment

FAQ from MCP Server for ReportServer Java Application

What are the prerequisites for running this server?

You need .NET 9.0 SDK or later, Java JDK 17 or later (for ReportServer), Docker (for Aspire containerized services), and an Ollama-compatible AI model (default: mistral-nemo:12b for chat, all-minilm for embeddings).

How do I troubleshoot ReportServer connection issues?

Verify that your Java ReportServer is running and accessible at the configured address in appsettings.json (default localhost:1099). Check connectivity from the ReportServerRPCClient and use the test-mcp-server.sh script to verify the connection.

Can I customize the AI model or ReportServer address?

Yes. In the appsettings.json file, you can change the Ollama model used for chat completions, configure the embedding model, and set the ReportServer address.

What is the current development status?

The project is under active development. Current focus areas include completing the ReportServer RPC interface, enhancing the MCP client connection, and enabling independent MCP server deployment with VS Code extension integration.

How do I add more report types or support other AI models?

Extend report capabilities by adding new methods to ReportServerRPCClient, creating corresponding functions in McpReportServer, and decorating them with the [McpServerTool] attribute. For other AI models, update RSChatApp.Web/Program.cs to use a different model provider and add the necessary packages.