MCP AI Server for Visual Studio

Other tools read your files. MCP AI Server understands your code.

The first and only Visual Studio extension that gives AI assistants access to the C# compiler (Roslyn) and the Visual Studio Debugger through the Model Context Protocol. 41 tools. 13 powered by Roslyn. 19 debugging tools (Preview). Semantic understanding, not text matching.

Install

Install from Visual Studio Marketplace

Or search for "MCP AI Server" in Visual Studio Extensions Manager.

See it in Action

Click to watch on YouTube

Why MCP AI Server?

AI coding tools (Claude Code, Codex CLI, Gemini CLI, OpenCode, Cursor, Copilot, Windsurf...) operate at the filesystem level — they read text, run grep, execute builds. They don't understand your code the way Visual Studio does.

MCP AI Server bridges this gap by exposing IntelliSense-level intelligence and the Visual Studio Debugger as MCP tools. Your AI assistant gets the same semantic understanding that powers F12 (Go to Definition), Shift+F12 (Find All References), safe refactoring — and now runtime debugging.

What becomes possible:

You ask	Without MCP AI Server	With MCP AI Server
"Find WhisperFactory"	grep returns 47 matches	`Class, Whisper.net, line 16` — one exact answer
"Rename ProcessDocument"	sed breaks ProcessDocumentAsync	Roslyn renames 23 call sites safely
"What implements IDocumentService?"	Impossible via grep	Full inheritance tree with interfaces
"What calls AuthenticateUser?"	Text matches, can't tell direction	Precise call graph: callers + callees
"Why is ProcessOrder returning null?"	Reads code, guesses	Sets breakpoint, inspects actual runtime values

22 Stable Tools

Semantic Navigation (Roslyn-powered)

Tool	Description
`FindSymbols`	Find classes, methods, properties by name — semantic, not text
`FindSymbolDefinition`	Go to definition (F12 equivalent)
`FindSymbolUsages`	Find all references, compiler-verified (Shift+F12 equivalent)
`GetSymbolAtLocation`	Identify the symbol at a specific line and column
`GetDocumentOutline`	Semantic structure: classes, methods, properties, fields

Code Understanding (Roslyn-powered)

Tool	Description
`GetInheritance`	Full type hierarchy: base types, derived types, interfaces
`GetMethodCallers`	Which methods call this method (call graph UP)
`GetMethodCalls`	Which methods this method calls (call graph DOWN)

Code Analysis (Roslyn-powered)

Tool	Description
`GetDiagnostics`	Compiler errors & warnings without building — Roslyn background analysis

Refactoring (Roslyn-powered)

Tool	Description
`RenameSymbol`	Safe rename across the entire solution — compiler-verified
`FormatDocument`	Visual Studio's native code formatter

Project & Build

Tool	Description
`ExecuteCommand`	Build or clean solution/project with structured diagnostics
`ExecuteAsyncTest`	Run tests asynchronously with real-time status
`GetSolutionTree`	Solution and project structure
`GetProjectReferences`	Project dependency graph
`LoadSolution`	Open a .sln/.slnx file — server stays on the same port
`TranslatePath`	Convert paths between Windows and WSL formats

Editor Integration

Tool	Description
`GetActiveFile`	Current file and cursor position
`GetSelection` / `CheckSelection`	Read active text selection
`GetLoggingStatus` / `SetLogLevel`	Extension diagnostics

19 Debugging Tools (Preview)

Your AI assistant can now debug your .NET code at runtime through the Visual Studio Debugger. Set breakpoints, step through code, inspect variables, attach to Docker containers and WSL processes.

Debug Control (10 tools)

Tool	Description
`debug_start`	Start debugging (F5). Fire-and-forget
`debug_stop`	Stop debugging session
`debug_get_mode`	Current mode: Design, Running, or Break
`debug_break`	Pause the running application
`debug_continue`	Resume execution
`debug_step`	Step over/into/out
`immediate_execute`	Execute expression with side effects
`debug_list_transports`	List transports (Default, Docker, WSL, SSH...)
`debug_list_processes`	List processes on a transport
`debug_attach`	Attach to a running process

Debug Inspection (5 tools)

Tool	Description
`debug_get_callstack`	Call stack of current thread
`debug_get_locals`	Local variables (tree-navigable)
`debug_evaluate`	Evaluate expression / drill into variable tree
`output_read`	Read VS Output window (Build, Debug, Tests)
`error_list_get`	Errors and warnings from VS Error List

Breakpoint Management (4 tools)

Tool	Description
`breakpoint_set`	Set breakpoint by file+line or function name
`breakpoint_remove`	Remove breakpoint
`breakpoint_list`	List all breakpoints
`exception_settings_set`	Configure break-on-exception

AI Debugging Guide

Complete reference for AI agents — all 19 tools, 10 workflows, Docker & WSL setup, polling patterns, and best practices.

Download AI Debugging Guide (.md) — add it to your AI's context for full debugging capabilities.

Compatible Clients

Works with any MCP-compatible AI tool:

CLI Agents:

Claude Code — Anthropic's terminal AI coding agent
Codex CLI — OpenAI's terminal coding agent
Gemini CLI — Google's open-source terminal agent
OpenCode — Open-source AI coding agent (45k+ GitHub stars)
Goose — Block's open-source AI agent
Aider — AI pair programming in terminal

Desktop & IDE:

Claude Desktop — Anthropic's desktop app
Cursor — AI-first code editor
Windsurf — Codeium's AI IDE
VS Code + Copilot — GitHub Copilot with MCP
Cline — VS Code extension
Continue — Open-source AI assistant

Any MCP client — open protocol, universal compatibility

Quick Start

Install from Visual Studio Marketplace
Open your .NET solution in Visual Studio
Configure port in MCP Server Settings (default: 3010)
Add to your MCP client:

json

"vs-mcp": {
  "type": "http",
  "url": "http://localhost:3010/sdk/"
}

Start asking your AI semantic questions about your code

MCP Client Configuration

AI Tool	Type	Config File
Claude Code	CLI	`CLAUDE.md`
Claude Desktop	App	`claude_desktop_config.json`
Cursor	IDE	`.cursor/rules/*.mdc` or `.cursorrules`
Windsurf	IDE	`.windsurfrules`
Cline	VS Code Extension	`.clinerules/` directory
VS Code + Copilot	IDE	`.github/copilot-instructions.md`
Continue	IDE Extension	`.continue/config.json`
Gemini CLI	CLI	`GEMINI.md`
OpenAI Codex CLI	CLI	`AGENTS.md`
Goose	CLI	`.goose/config.yaml`

Any tool supporting Model Context Protocol will work.

Configure AI Preferences (Recommended)

Add these instructions to your project's AI config file (see table above) to ensure your AI automatically prefers MCP tools:

markdown

## MCP Tools - ALWAYS PREFER

When `mcp__vs-mcp__*` tools are available, ALWAYS use them instead of Grep/Glob/LS:

| Instead of | Use |
|------------|-----|
| `Grep` for symbols | `FindSymbols`, `FindSymbolUsages` |
| `LS` to explore projects | `GetSolutionTree` |
| Reading files to find code | `FindSymbolDefinition` then `Read` |
| Searching for method calls | `GetMethodCallers`, `GetMethodCalls` |

**Why?** MCP tools use Roslyn semantic analysis - 10x faster, 90% fewer tokens.

Multi-Solution: Understand Your Dependencies

Need the AI to understand a library you depend on? Clone the source from GitHub, open it in a second Visual Studio — each instance runs its own MCP server on a configurable port.

Code

Your project                          → port 3010
Library source (cloned from GitHub)   → port 3011
Framework source                      → port 3012

Your AI connects to all of them. It can trace calls, find usage patterns, and understand inheritance across your code and library code.

MCP client configuration — three options depending on your client:

Clients with native HTTP support (Claude Desktop, Claude Code):

config.json

{
  "mcpServers": {
    "vs-mcp": {
      "type": "http",
      "url": "http://localhost:3010/sdk/"
    },
    "vs-mcp-whisper": {
      "type": "http",
      "url": "http://localhost:3011/sdk/"
    }
  }
}

Clients without HTTP support (via mcp-remote proxy):

config.json

{
  "mcpServers": {
    "vs-mcp": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "http://localhost:3010/sdk/"]
    },
    "vs-mcp-whisper": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "http://localhost:3011/sdk/"]
    }
  }
}

Codex CLI (TOML):

toml

[mcp_servers.vs-mcp]
type = "stdio"
command = "npx"
args = ["mcp-remote", "http://localhost:3010/sdk/"]

[mcp_servers.vs-mcp-whisper]
type = "stdio"
command = "npx"
args = ["mcp-remote", "http://localhost:3011/sdk/"]

Settings

Tools → MCP Server Settings

VS-MCP Settings

Setting	Default	Description
Port	3001	Server port (configurable per VS instance)
Path Format	WSL	Output as `/mnt/c/...` or `C:\...`
Tools	All enabled	Enable/disable individual tool groups

Changes apply on next server start.

vs. Other Approaches

Approach	Symbol Search	Inheritance	Call Graph	Safe Rename	Debugging
MCP AI Server (Roslyn + Debugger)	Semantic	Full tree	Callers + Callees	Compiler-verified	Breakpoints + Step + Inspect
AI Agent (grep/fs)	Text match	No	No	Text replace	No
Other MCP servers	No	No	No	No	No

Requirements

Visual Studio 2022 (17.13+) or Visual Studio 2026
Windows (amd64 or arm64)
Any MCP-compatible AI tool

Issues & Feature Requests

Found a bug or have an idea? Open an issue!

About

0ics srl — Italian software company specializing in AI-powered development tools. Part of the example4.ai ecosystem.

Built by Ladislav Sopko — 30 years of software development, from assembler to enterprise .NET.

MCP AI Server: Because your AI deserves the same intelligence as your IDE.