McpServerManager is a cross-platform Avalonia UI desktop application that visualizes and analyzes session logs and request data produced by AI coding assistants such as GitHub Copilot and Cursor. It provides a unified dashboard that aggregates logs from different providers, a tree-style navigation pane that indexes log files from a monitored directory, and a detailed request view that exposes interpretation, metadata, context, and automated actions taken by the AI tools. Use the app to review prompt history, inspect the exact context sent to the LLM, and track actions like file creation or edits in a structured grid. The built-in JSON inspector and Markdown renderer (Markdig.Avalonia) help you debug data or render log documentation directly within the UI.

Prerequisites:

• .NET 8 SDK or later
• Git

Install steps:

1. Clone the repository: git clone https://github.com/sharpninja/McpServerManager.git
2. Build or run the project directly: cd McpServerManager

   From repo root, using the specified project path

   dotnet run --project src/McpServerManager

Optional:

• If you want to publish a standalone app, use: dotnet publish --project src/McpServerManager -c Release -r <RID> -p:PublishSingleFile=true (Choose RID for your platform, e.g., win-x64, linux-x64, osx-x64)

Tips and notes:

• The app monitors a target logs directory (configured in appsettings.config) and indexes JSON/Markdown log files into a navigation tree for quick access.
• Session logs are loaded from Mcp.BaseUrl /mcpserver/sessionlog when used in a networked setup; Paths.* entries point to documents and CSS fallback content used by the Markdown renderer.
• Environment variables referenced by Paths.HtmlCacheDirectory can include system variables like %TEMP% on Windows.
• If you encounter rendering issues with Markdown, ensure Markdig.Avalonia is correctly referenced and that the log data includes valid Markdown blocks.
• For debugging, use the Original JSON expander in the request view to inspect the raw data fields that were captured.