10x MCP Server

Setup

The MCP server can be found here:

C:\Program Files\PureDevSoftware\10x\MCPServer\10xMCPServer.exe

Follow the instructions for installing the MCP server exe with your AI agent.

Setup for Claude Code

Setup for Cursor

Functionality

Once the 10x MCP server is installed the AI agent will have full access to all 10x commands and python scripting.

Below is a list of some of the MCP tools available.

list_editors	List all 10x Editor instances that can be connected to
connect_to	Connect to a specific 10x Editor instance
read_current_buffer	Read the currently open file
write_to_buffer	Replace the currently open file
edit_range	Replace a specific range in the currently open file
get_selection	Get the currently selected text
get_current_file	Get the file path of the currently open file
get_open_files	Get a list of all open files
open_file	Open a specific file
get_cursor_position	Get the current cursor position
set_cursor_position	Set the current cursor position
run_last_build	Run the last build
debug_project	Start the debugger
build_project	Build the currently active project
rebuild_project	Rebuild the currently active project
get_project_files	Get a list of all files in the project
save_current_buffer	Save the currently open file
search_in_file	Find text in the currently open file
search_in_project	Use Workspace Search to search for text in the workspace
find_file_by_name	Use Workspace Search to find a file in the workspace by name
get_build_output	Get the build output
get_build_results	Get the build results (warnings and errors)
get_build_status	Return if the build is still running
get_file_diagnostics	Get warnings and errors for the current file (without doing a build)
execute_command	Execute a 10x command or python statement
list_commands	List all 10x commands

For a full list of tool use: /mcp and select 10x Editor, then View Tools

Usage

Below are some recommendations on how to use 10x with AI agents:

• Prompt: "What does this file do?"
  AI agent reads the current buffer and gives a high-level summary, useful when jumping into unfamiliar code
• Prompt: "Describe the current function"
  AI agent will look at the currently open file and the cursor position and analyise the function.
• Prompt: "Add a documentation comment to the current function"
• Prompt: "Connect to MyTestApp"
  AI agent will find a 10x instance with MyTestAppp workspace open and connect to that instance
• Prompt: "Find all usages of MyFunction in the project"
• Prompt: "What files are open?"
  AI agent uses get_open_files to give an overview of what you're working with
• Prompt: "Refactor the current function to remove code duplication"
• Prompt: "Build the project and fix any errors"
  AI agent runs build_project, reads get_build_results, opens the relevant files, and iterates until clean
• Prompt: "The last build failed, what went wrong?"
  AI agent reads get_build_output and explains the errors
• Prompt: "Start the debugger"
• Prompt: "Write a test for the current function"
  AI agent reads the function and generates a test case alongside it
• Prompt: "Find all TODO comments in the project and list them"