MCP Catalogs
Home

dap_mcp

by KashunCheng·40·Score 42

DAP-MCP enables LLM interaction with debuggers through standardized MCP protocol, offering debugging tools like breakpoints and execution control.

developer-toolsai-llmops-infra
6
Forks
1
Open issues
8 mo ago
Last commit
2d ago
Indexed

Overview

dap-mcp is a Python implementation of the Model Context Protocol that bridges large language models with debug adapters. It provides a set of tools for controlling debugging sessions, including setting and removing breakpoints, controlling execution flow (continue, step in/out/next), evaluating expressions, changing stack frames, and viewing source code. The server supports multiple debug adapters like debugpy and lldb through a flexible JSON configuration system, making it adaptable to various debugging environments.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:AI-assisted debugging where language models can set breakpoints and control execution flow
you:Automated debugging workflows triggered by LLMs based on code analysis
you:Remote debugging through LLM interfaces when direct debugger access is limited
you:What debuggers are supported by dap-mcp?
you:How do I add support for a new debugger?

When to choose this

Choose dap-mcp when you need AI assistance for debugging workflows and your team uses Python (debugpy) or lldb debuggers.

When NOT to choose this

Avoid if you use other debuggers not yet supported, or need a more comprehensive IDE integration beyond MCP protocol support.

Tools this server exposes

12 tools extracted from the README
  • launch

    Launch the debuggee program.

  • set_breakpoint

    Set a breakpoint at a specified file and line (with an optional condition).

  • remove_breakpoint

    Remove a breakpoint from a specified file and line.

  • list_all_breakpoints

    List all breakpoints currently set in the debugger.

  • continue_execution

    Continue program execution after hitting a breakpoint.

  • step_in

    Step into a function call.

  • step_out

    Step out of the current function.

  • next

    Step over to the next line of code.

  • evaluate

    Evaluate an expression in the current debugging context.

  • change_frame

    Switch to a different stack frame.

  • view_file_around_line

    View source code around a specified line.

  • terminate

    Terminate the debugging session.

Comparable tools

vscode-mcpdebugpylldb-dap

Installation

Installation

  1. Install the dap-mcp package:

``bash pip install dap-mcp ``

  1. Create a configuration file (e.g., config.json) with your debugger settings
  1. Run the server:

``bash python -m dap_mcp --config config.json # or with uv uvx dap-mcp@latest --config config.json ``

Claude Desktop Configuration

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "dap-debugger": {
      "command": "python",
      "args": ["-m", "dap_mcp"],
      "env": {
        "DAP_MCP_CONFIG": "/path/to/config.json"
      }
    }
  }
}

FAQ

What debuggers are supported by dap-mcp?
Currently, dap-mcp supports debugpy and lldb through the Debug Adapter Protocol. The architecture is designed to be extensible, allowing for additional debugger support by creating new configuration classes.
How do I add support for a new debugger?
To add support for a new DAP server, create a new configuration class in `dap_mcp/config.py` that extends the base `DAPConfig` class. Define a unique type value and any additional settings specific to that debugger, then update the union type for debugger-specific configurations.

Compare dap_mcp with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.