home / mcp / binlog mcp server
A simple MCP server that analyzes MSBuild binlogs to identify expensive targets, project/build times, evaluations, and embedded file access, with fast, cached queries.
Configuration
View docs{
"mcpServers": {
"baronfel-mcp-msbuild-host": {
"command": "<repo root>\\binlog.mcp\\bin\\Debug\\net9.0\\binlog.mcp.exe",
"args": []
}
}
}Binlog MCP Server exposes tools and prompts for analyzing MSBuild binlogs to any MCP client. It helps you identify expensive targets, analyze project build times, inspect evaluations, access embedded source files, and benefit from intelligent caching for fast queries across large binlogs.
Connect to the server from your MCP client to begin analyzing MSBuild binlogs. You can run the server locally and then point your client to the local endpoint, or use an already deployed remote MCP URL if available. The server exposes tools for target analysis, project analysis, evaluation analysis, and file access, which you can invoke via standard MCP client commands. For quick setup, you can debug locally with an inspector that talks to the MCP endpoint.
Prerequisites: you need the .NET SDK installed to build the server, and you should have Node.js available if you plan to use the quick inspection workflow.
Step 1: build the MCP server locally
Step 2: configure your MCP client (Claude or VSCode) to connect to the server endpoint
Step 3: launch your server app and start querying MSBuild binlogs
{
"mcpServers": {
"msbuild": {
"command": "<your repo root>\\binlog.mcp\\bin\\Debug\\net9.0\\binlog.mcp.exe"
}
}
}To locally debug and inspect the MCP server, you can run the Model Context Protocol inspector against the built server binary.
npx @modelcontextprotocol/inspector ./bin/Debug/net9.0/msbuild.mcpIdentify expensive targets, search for specific targets across projects, and analyze target execution times.
Calculate project build times, find the most expensive projects, and analyze all targets in a project at once.
List project evaluations, inspect global properties, and identify potential overbuilding issues.
List and retrieve source files embedded in binary logs.
Intelligent caching to speed up queries on large binlogs.