# hechtcarmel/jetbrains-debugger-mcp-plugin [Health: Active]

**Category:** 💻 Developer Tools  
**Repository:** https://github.com/hechtcarmel/jetbrains-debugger-mcp-plugin  
**GitHub Stars:** 105  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/hechtcarmel-jetbrains-debugger-mcp-plugin

## Description
A JetBrains IDE plugin that exposes an MCP server, giving AI coding assistants full programmatic control over the debugger.

## Tools
Capabilities this server exposes over MCP:

- **list_run_configurations** — List all available run configurations in the project
- **execute_run_configuration** — Execute a run configuration in debug or run mode
- **list_debug_sessions** — List all active debug sessions with state and metadata
- **start_debug_session** — Start a new debug session for a run configuration
- **stop_debug_session** — Stop/terminate a debug session
- **get_debug_session_status** — Get comprehensive status (variables, stack, source) in one call
- **list_breakpoints** — List all breakpoints with optional filtering
- **set_breakpoint** — Set a line breakpoint with condition, log message, suspend policy
- **remove_breakpoint** — Remove a breakpoint by ID or location
- **resume_execution** — Resume paused execution
- **pause_execution** — Pause running execution
- **step_over** — Step over to next line (without entering methods)
- **step_into** — Step into method calls
- **step_out** — Step out of current method
- **run_to_line** — Continue execution until a specific line
- **wait_for_pause** — Block until the session pauses (breakpoint, step, manual) and return full status
- **get_stack_trace** — Get current call stack with file/line/method info
- **select_stack_frame** — Change debugger context to a different stack frame
- **list_threads** — List all threads with state information
- **get_variables** — Get all variables visible in current stack frame
- **set_variable** — Modify a variable's value during debugging
- **get_source_context** — Get source code around current execution point
- **evaluate_expression** — Evaluate an expression or code fragment in debug context

## Claude Desktop Quick Installation
Heuristic fallback — verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low):

```json
"mcpServers": {
  "jetbrains-debugger-mcp-plugin": {
    "command": "npx",
    "args": ["-y","hechtcarmel-jetbrains-debugger-mcp-plugin"]
  }
}
```

## Documentation

## What hechtcarmel/jetbrains-debugger-mcp-plugin MCP server does

The hechtcarmel/jetbrains-debugger-mcp-plugin MCP server adds an MCP interface to the debugger supplied by a JetBrains IDE. It lets an MCP client work with run configurations and active debug sessions instead of relying on manual IDE interaction.

The available operations cover the debugging lifecycle: discovering run configurations, starting or stopping sessions, checking session state, waiting for execution to pause, and resuming or pausing a process. Multiple concurrent debug sessions are supported. A status request can include variables, stack information, and source context in one response.

## How it works

Install the plugin into a JetBrains IDE, open a project, and the MCP server starts automatically on an IDE-specific port. The server exposes a streamable HTTP endpoint using this path pattern:

`http://127.0.0.1:{PORT}/debugger-mcp/streamable-http`

The port and host binding are configurable in the IDE settings under Tools > Debugger MCP Server. IntelliJ IDEA uses port 29190 by default; the README lists separate defaults for several other IDEs. The tool window can display server activity and copy client configuration.

The hechtcarmel/jetbrains-debugger-mcp-plugin MCP server is designed for clients that can connect to an HTTP MCP endpoint. The supplied configuration examples cover Claude Code, Codex CLI, Gemini CLI, Cursor, Windsurf, and generic MCP clients. A bundled companion skill provides guidance for AI agents using the debugger tools and can be installed into `.claude/skills/` or exported as a skill or ZIP file.

## Setup and configuration

Install “Debugger MCP Server” through the JetBrains plugin marketplace, use the Marketplace installation page, or download a plugin version and install it from disk. Restart the IDE, open the target project, then check Settings or Preferences > Tools > Debugger MCP Server for the active port.

Client setup consists of adding the endpoint to the MCP client. Cursor uses `url`, while Windsurf uses `serverUrl`; both examples point to the streamable HTTP endpoint. The IDE tool window has an “Install on Coding Agents” action that can run setup for Claude Code or Codex CLI, copy configuration for supported clients, and show generic streamable HTTP or legacy SSE configuration.

## Tools and capabilities

The hechtcarmel/jetbrains-debugger-mcp-plugin MCP server provides tools for:

- Listing, running, and debugging project run configurations.
- Listing sessions, reading status, starting sessions, and terminating sessions.
- Creating and removing line breakpoints, including conditions, log messages, and suspend policies.
- Resuming, pausing, stepping over, stepping into, stepping out, and running to a line.
- Waiting for a pause caused by a breakpoint, step, or manual action.
- Reading stack traces, selecting frames, listing threads, and retrieving source context.
- Inspecting visible variables and changing variable values during a debug session.
- Evaluating expressions and multi-line code fragments with configurable evaluation safety controls.

## Limitations and notes

IDE support is not uniform. CI verification covers IntelliJ IDEA, PyCharm, WebStorm, and GoLand. The README identifies RustRover, Android Studio, and PhpStorm as community-tested, while RubyMine, CLion, and DataGrip are listed as untested but potentially compatible. The plugin targets JetBrains IDEs that support XDebugger.

The server endpoint is configured per IDE and may use a different port for each product. The provided material does not describe authentication, remote hosting, or a standalone command-line server; operation depends on the plugin running inside the IDE. Expression evaluation has configurable safety modes, including unrestricted evaluation, a default risky-operation blocklist, and conservative read-only checks with optional custom regular-expression rules.

_Full upstream README: https://allmcps.com/mcp/hechtcarmel-jetbrains-debugger-mcp-plugin/readme_