hechtcarmel/jetbrains-index-mcp-plugin

What it solves

This plugin bridges the gap between AI coding assistants (like Claude, Cursor, and Windsurf) and the deep semantic understanding of JetBrains IDEs. While most AI tools rely on simple text search or basic file reading, this project allows them to use the IDE's actual index, AST (Abstract Syntax Tree), and refactoring engines to navigate and modify code with high precision.

How it works

It implements a Model Context Protocol (MCP) server directly within the JetBrains IDE. This server exposes a wide array of tools that the AI assistant can call via HTTP. When an AI requests a specific action—such as finding all references to a symbol or renaming a variable—the plugin executes that command using the IDE's internal APIs and returns the structured result to the AI.

Who it’s for

Developers who use JetBrains IDEs (IntelliJ IDEA, PyCharm, WebStorm, GoLand, etc.) and want their AI coding agents to have professional-grade code intelligence, safe refactoring capabilities, and the ability to handle complex, multi-language projects.

Highlights

• Deep Semantic Intelligence: Provides tools for finding definitions, references, type hierarchies, and call hierarchies across multiple languages.
• Safe Refactoring: Enables AI to perform safe renames and file moves that automatically update all related references.
• Multi-Language Support: Works across Java, Kotlin, Python, JavaScript, TypeScript, Go, PHP, and Rust.
• Project Lifecycle Management: Includes an opt-in system to sleep or wake projects to save memory when working with many projects simultaneously.
• Broad IDE Compatibility: Tested on most major JetBrains IDEs, including Android Studio and RustRover.