MXterm Pro

Developer notes

Build better inside the terminal.

Guides on running AI coding agents, managing SSH/SFTP servers, and getting more out of MXterm Pro.

Featured guide

MobaXterm for Mac: How to Run It, or What to Use Instead on macOS

MobaXterm doesn't run on macOS and Mobatek hasn't announced a port. Here's why, what Wine/CrossOver actually get you, and how MXterm Pro brings SSH, SFTP, and AI agent panes natively to Mac and Windows.

Read guide →

AI Agent Terminal & Control Room

View all →
AI CONTROL ROOM

Orchestrating Parallel AI Agents in Split Terminal Panes

A step-by-step guide to running Claude Code and Codex concurrently in split terminal panes, with git worktree isolation so neither agent stomps on the other's branch.

Read →
COMMAND RECIPES

How Command Recipes Works: Ready-Made Linux Commands, Matched by Keyword

Press Cmd+J, type a task in English or Vietnamese, and get the matching ready-made Linux command with a one-line description: keyword-matched inside the app, no AI model, nothing sent anywhere.

Read →
CURSOR

Running Cursor Agent in a Multi-Pane Terminal on Mac

A real walkthrough of running Cursor's CLI agent in MXterm Pro: setup steps, how live status detection actually works, git worktree isolation, and the one thing session resume still can't do automatically.

Read →
OLLAMA

Run Ollama in a Terminal Alongside SSH Sessions

A split-pane workflow for local-model development: run Ollama in one terminal pane, SSH into remote servers in another, and reach either one from the other without leaving your keyboard.

Read →
DOCKER

Docker Build Terminal on Mac: Builds Alongside SSH Deploys

Keep a docker build running in one pane and an SSH session to your deploy target open in the next: split with a keystroke, no window switching, no separate SFTP client for the files the container needs.

Read →
GIT WORKTREE

Isolating Parallel AI Agents with Git Worktrees

Two or more coding agents in one repo will eventually collide on the working tree. Here's exactly how MXterm Pro's worktree-per-agent isolation stops that: and where it doesn't.

Read →
AI CONTROL ROOM

Control Room (formerly War Room): Running Claude, Codex & Gemini in Parallel

Real mechanism, not marketing copy: how MXterm Pro's Control Room lets Claude Code, Codex, and Gemini CLI message each other through MCP tools, when it earns its keep, and how it compares to relaying between panes by hand.

Read →
TOKEN METER

Claude Code and Codex Token Usage: What the Token Meter Reads

What MXterm Pro's Token Meter shows for Claude Code, Codex, Gemini CLI and omp, why cache reads get their own line, and when it shows a percentage instead of a token count.

Read →
PROVIDER BALANCE

Check AI API Credit Balance Without Opening Five Dashboards

DeepSeek and Moonshot report money, Kimi, MiniMax and Z.ai report a plan quota, OpenRouter reports either. What each endpoint gives you, and how to read all six from one place.

Read →
TOKEN METER

Usage, Plan Limit, and Balance: Three Numbers People Keep Confusing

Tokens spent, plan percentage, and account balance are three separate measurements from three sources. When each one matters, and how each one misleads.

Read →

Ask AI about MXterm Pro

Open a live answer in ChatGPT, Claude or Perplexity, based on this page. The assistant writes it, not us, so check the sources it cites.

Question sentWhich MXterm Pro guides cover running AI coding agents and managing SSH servers?