LLDB MCP Server

Instructions

Implementation Reference

• lldb_mcp_server.py:661-708 (handler)

  The async handler function implementing the core logic of the lldb_examine_variables tool. It constructs LLDB commands to set a breakpoint, run the program, examine variables, and formats the output as Markdown or JSON.
  async def lldb_examine_variables(params: ExamineVariablesInput) -> str: """Examine local variables and arguments at a breakpoint. Runs the program until the specified breakpoint, then displays the values of local variables and function arguments. Args: params: ExamineVariablesInput with executable, breakpoint, and optional variable names Returns: str: Variable values at the breakpoint """ commands = [ f"target create {params.executable}", f"breakpoint set --name {params.breakpoint}", "run" + (" " + " ".join(params.args) if params.args else ""), ] if params.variables: for var in params.variables: commands.append(f"frame variable {var}") else: commands.append("frame variable") commands.append("quit") result = _run_lldb_script(commands) if params.response_format == ResponseFormat.JSON: return json.dumps( { "success": result["success"], "breakpoint": params.breakpoint, "output": result["output"], "error": result.get("error"), }, indent=2, ) lines = [ f"## Variables at `{params.breakpoint}`", "", "```", result["output"].strip() if result["success"] else result.get("error", "Unknown error"), "```", ] return "\n".join(lines)
• lldb_mcp_server.py:225-241 (schema)

  Pydantic BaseModel defining the input schema (parameters) for the lldb_examine_variables tool.
  class ExamineVariablesInput(BaseModel): """Input for examining variables.""" model_config = ConfigDict(str_strip_whitespace=True) executable: str = Field(..., description="Path to the executable", min_length=1) breakpoint: str = Field(..., description="Breakpoint location to stop at", min_length=1) variables: list[str] | None = Field( default=None, description="Specific variable names to examine (if None, shows all locals)" ) args: list[str] | None = Field( default=None, description="Command-line arguments to pass to the program" ) response_format: ResponseFormat = Field( default=ResponseFormat.MARKDOWN, description="Output format" )
• lldb_mcp_server.py:651-660 (registration)

  The @mcp.tool decorator that registers the lldb_examine_variables function as an MCP tool with specified name and annotations.
  @mcp.tool( name="lldb_examine_variables", annotations={ "title": "Examine Variables", "readOnlyHint": True, "destructiveHint": False, "idempotentHint": True, "openWorldHint": False, }, )
• lldb_mcp_server.py:118-156 (helper)

  Helper utility function used by the tool (and others) to execute a list of LLDB commands in batch mode and return structured results.
  def _run_lldb_script( commands: list[str], target: str | None = None, working_dir: str | None = None, timeout: int = 60, ) -> dict[str, Any]: """ Execute multiple LLDB commands in sequence. """ cmd = [LLDB_EXECUTABLE] if target: cmd.extend(["--file", target]) cmd.append("--batch") for command in commands: cmd.extend(["-o", command]) try: result = subprocess.run( cmd, capture_output=True, text=True, timeout=timeout, cwd=working_dir or os.getcwd() ) return { "success": result.returncode == 0, "output": result.stdout, "error": result.stderr if result.returncode != 0 else None, "return_code": result.returncode, } except subprocess.TimeoutExpired: return { "success": False, "output": "", "error": f"Commands timed out after {timeout} seconds", "return_code": -1, } except Exception as e: return {"success": False, "output": "", "error": str(e), "return_code": -1}