MCP Directory

title: "MCP Setup in JetBrains IDEs (PyCharm, IntelliJ, WebStorm)" description: "Configure Model Context Protocol servers in JetBrains IDEs including PyCharm, IntelliJ IDEA, and WebStorm." slug: "jetbrains-integration" category: "ide" updatedAt: "2025-09-21T00:00:00.000Z" faqs:

  • q: "Which JetBrains IDEs support MCP integration?" a: "All major JetBrains IDEs support MCP through plugins: IntelliJ IDEA, PyCharm, WebStorm, PhpStorm, GoLand, and others."
  • q: "Do I need different MCP configurations for different JetBrains IDEs?" a: "The MCP configuration is largely the same across JetBrains IDEs, with minor differences based on language-specific features."
IDEs
MCP SDK v2.1.0
Updated Sep 21, 20255 min read
jetbrains
pycharm
intellij
webstorm
integration

MCP Setup in JetBrains IDEs (PyCharm, IntelliJ, WebStorm)

Overview

JetBrains IDEs offer excellent support for Model Context Protocol through plugins and built-in integrations. This guide covers setup for IntelliJ IDEA, PyCharm, WebStorm, and other JetBrains IDEs.

Prerequisites

  • JetBrains IDE (2023.1 or later)
  • JetBrains Toolbox (recommended)
  • Node.js 18+ or Python 3.8+
  • JetBrains account (for plugin access)

Installation

Step 1: Install MCP Plugin

  1. Open your JetBrains IDE
  2. Go to File → Settings (or Preferences on macOS)
  3. Navigate to Plugins
  4. Search for "Model Context Protocol"
  5. Install the MCP Integration plugin
  6. Restart the IDE

Step 2: Install MCP Servers

# For Java/Kotlin projects (IntelliJ IDEA)
npm install -g @modelcontextprotocol/server-filesystem
npm install -g @modelcontextprotocol/server-git

# For Python projects (PyCharm)
pip install mcp-server-python
pip install mcp-server-database

# For JavaScript/TypeScript projects (WebStorm)
npm install -g @modelcontextprotocol/server-nodejs
npm install -g @modelcontextprotocol/server-webpack

IDE-Specific Configuration

IntelliJ IDEA

File → Settings → Tools → MCP Integration

<mcp-configuration>
  <servers>
    <server name="filesystem" enabled="true">
      <command>mcp-server-filesystem</command>
      <args>
        <arg>--root</arg>
        <arg>${PROJECT_DIR}</arg>
      </args>
      <env>
        <var name="JAVA_HOME" value="${JDK_HOME}" />
      </env>
    </server>
    
    <server name="maven" enabled="true">
      <command>mcp-server-maven</command>
      <args>
        <arg>--pom</arg>
        <arg>${PROJECT_DIR}/pom.xml</arg>
      </args>
    </server>
    
    <server name="git" enabled="true">
      <command>mcp-server-git</command>
      <args>
        <arg>--repo</arg>
        <arg>${PROJECT_DIR}</arg>
      </args>
    </server>
  </servers>
</mcp-configuration>

PyCharm

File → Settings → Tools → MCP Integration

<mcp-configuration>
  <servers>
    <server name="python" enabled="true">
      <command>python</command>
      <args>
        <arg>-m</arg>
        <arg>mcp_server_python</arg>
        <arg>--project</arg>
        <arg>${PROJECT_DIR}</arg>
      </args>
      <env>
        <var name="PYTHONPATH" value="${PROJECT_DIR}" />
        <var name="VIRTUAL_ENV" value="${VIRTUAL_ENV}" />
      </env>
    </server>
    
    <server name="database" enabled="true">
      <command>mcp-server-database</command>
      <env>
        <var name="DATABASE_URL" value="${DATABASE_URL}" />
      </env>
    </server>
    
    <server name="django" enabled="true">
      <command>python</command>
      <args>
        <arg>-m</arg>
        <arg>mcp_server_django</arg>
        <arg>--settings</arg>
        <arg>${PROJECT_DIR}/settings.py</arg>
      </args>
    </server>
  </servers>
</mcp-configuration>

WebStorm

File → Settings → Tools → MCP Integration

<mcp-configuration>
  <servers>
    <server name="nodejs" enabled="true">
      <command>mcp-server-nodejs</command>
      <args>
        <arg>--project</arg>
        <arg>${PROJECT_DIR}</arg>
      </args>
      <env>
        <var name="NODE_ENV" value="development" />
        <var name="NPM_CONFIG_PREFIX" value="${NPM_PREFIX}" />
      </env>
    </server>
    
    <server name="webpack" enabled="true">
      <command>mcp-server-webpack</command>
      <args>
        <arg>--config</arg>
        <arg>${PROJECT_DIR}/webpack.config.js</arg>
      </args>
    </server>
    
    <server name="typescript" enabled="true">
      <command>mcp-server-typescript</command>
      <args>
        <arg>--tsconfig</arg>
        <arg>${PROJECT_DIR}/tsconfig.json</arg>
      </args>
    </server>
  </servers>
</mcp-configuration>

Usage Features

Code Completion

MCP servers enhance IntelliSense and code completion:

// In IntelliJ IDEA
Files.readAllLines(
// MCP suggests files from filesystem server

// Database queries in PyCharm
cursor.execute(
// MCP suggests tables and columns from database server

AI Assistant Integration

JetBrains AI Assistant uses MCP context:

// Ask JetBrains AI Assistant:
"Refactor this class using filesystem MCP context"
"Generate tests based on the current project structure"
"Create a database migration using the database MCP schema"

Tool Windows

Access MCP features through dedicated tool windows:

  • MCP Servers: View and manage running servers
  • MCP Logs: Monitor server activity
  • MCP Tools: Quick access to MCP functions

Language-Specific Features

Java/Kotlin (IntelliJ IDEA)

<server name="spring" enabled="true">
  <command>mcp-server-spring</command>
  <args>
    <arg>--project</arg>
    <arg>${PROJECT_DIR}</arg>
    <arg>--profile</arg>
    <arg>dev</arg>
  </args>
  <env>
    <var name="SPRING_PROFILES_ACTIVE" value="development" />
  </env>
</server>

Python (PyCharm)

<server name="conda" enabled="true">
  <command>mcp-server-conda</command>
  <env>
    <var name="CONDA_DEFAULT_ENV" value="${CONDA_ENV}" />
  </env>
</server>

JavaScript/TypeScript (WebStorm)

<server name="react" enabled="true">
  <command>mcp-server-react</command>
  <args>
    <arg>--src</arg>
    <arg>${PROJECT_DIR}/src</arg>
  </args>
</server>

Advanced Configuration

Project-Level Configuration

Create .idea/mcp.xml in your project:

<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
  <component name="MCPConfiguration">
    <servers>
      <server name="project-specific">
        <command>node</command>
        <args>
          <arg>scripts/mcp-server.js</arg>
        </args>
        <workingDirectory>$PROJECT_DIR$</workingDirectory>
      </server>
    </servers>
  </component>
</project>

Environment Variables

Configure environment variables in Run/Debug Configurations:

<configuration>
  <envs>
    <env name="MCP_LOG_LEVEL" value="debug" />
    <env name="DATABASE_URL" value="postgresql://localhost:5432/mydb" />
    <env name="GITHUB_TOKEN" value="${GITHUB_TOKEN}" />
  </envs>
</configuration>

Debugging MCP Servers

Debug Configuration

  1. Go to Run → Edit Configurations
  2. Add new Node.js or Python configuration
  3. Set JavaScript file or Script path to your MCP server
  4. Add breakpoints in your MCP server code
  5. Start debugging

Log Monitoring

View MCP logs in the MCP Logs tool window:

[INFO] MCP Server started: filesystem
[DEBUG] Tool called: list_files
[ERROR] Connection failed: database server

Network Inspection

Use JetBrains' built-in network inspector:

  1. Tools → MCP → Network Inspector
  2. Monitor MCP protocol messages
  3. Analyze request/response patterns

Troubleshooting

Plugin Not Loading

  1. Check JetBrains IDE version compatibility
  2. Verify plugin installation in Settings → Plugins
  3. Restart IDE after installation
  4. Check IDE logs: Help → Show Log in Explorer

Server Connection Issues

# Test server manually
mcp-server-filesystem --test

# Check Java classpath (IntelliJ)
echo $CLASSPATH

# Check Python path (PyCharm)
python -c "import sys; print(sys.path)"

# Check Node.js path (WebStorm)
which node
npm config get prefix

Performance Issues

  1. Reduce MCP server count: Only enable needed servers
  2. Increase memory: Help → Edit Custom VM Options 
    -Xmx4g
-XX:ReservedCodeCacheSize=512m
  3. Optimize indexing: Settings → Build → Compiler → Exclude from indexing

Best Practices

Configuration Management

  • Use project-specific configurations
  • Store sensitive data in environment variables
  • Version control MCP configurations (.idea/mcp.xml)
  • Document custom MCP setups in README

Development Workflow

// Good: Use MCP context in code generation
// Ask AI: "Generate a REST controller using the database MCP schema"

// Better: Combine multiple MCPs
// Ask AI: "Create integration tests using filesystem and database MCPs"

Performance Optimization

  • Enable only necessary MCP servers
  • Use MCP server health checks
  • Monitor resource usage
  • Implement proper error handling

Integration Examples

Spring Boot (IntelliJ IDEA)

<server name="spring-boot" enabled="true">
  <command>mcp-server-spring-boot</command>
  <args>
    <arg>--application-properties</arg>
    <arg>${PROJECT_DIR}/src/main/resources/application.properties</arg>
  </args>
</server>

Django (PyCharm)

<server name="django-models" enabled="true">
  <command>python</command>
  <args>
    <arg>manage.py</arg>
    <arg>mcp_server</arg>
  </args>
  <workingDirectory>${PROJECT_DIR}</workingDirectory>
</server>

React (WebStorm)

<server name="react-components" enabled="true">
  <command>npx</command>
  <args>
    <arg>react-mcp-server</arg>
    <arg>--src</arg>
    <arg>src/components</arg>
  </args>
</server>

Related Guides

FAQ

Which JetBrains IDEs support MCP integration?

All major JetBrains IDEs support MCP through plugins: IntelliJ IDEA, PyCharm, WebStorm, PhpStorm, GoLand, CLion, Rider, and others. The plugin works consistently across the JetBrains platform.

Do I need different MCP configurations for different JetBrains IDEs?

The core MCP configuration is the same across JetBrains IDEs. However, you may want language-specific MCP servers (e.g., Spring for Java, Django for Python, React for JavaScript).

Can I share MCP configurations across team members?

Yes, commit the .idea/mcp.xml file to version control. Use environment variables for sensitive data and document setup requirements in your project README.

Was this guide helpful?

Last updated: September 21, 2025

Edit this page: jetbrains-integration/page.mdx