Powered by Claude Code  ·  Codex  ·  macOS native

Go get a coffee.
Your features
are building themselves.

For people who are done babysitting their agents across a dozen tabs. Works with Claude Code and Codex. Give it the state of the work. Walk away. Come back to something clear enough to trust.

Get TerminalAI $9  → See how it works
TerminalAI dashboard. Choose your next move.
BA
AH
SR
ML
Used by creative developers, directors who code, and solo operators running too many projects at once
Brandon Aviram
Why I built this

I was sitting down with Claude feeling like a monkey that has to keep pressing enter and watch things. Draining, physically and emotionally. I had 12 windows minimized in various states, copying context from one dead chat to the next. So I built the version where I could leave. Now I give it the hard problem, put down the laptop, and go to dinner. It’s done when I get back.

Brandon Aviram Director, developer, solo operator
Under the hood

Claude Code.
With a cockpit.

Terminal AI doesn't reinvent the engine. It runs Claude Code, Anthropic's agentic coding CLI, underneath everything. You're not betting on something unproven. You get the full power of Claude Code plus the interface and workflow layer that makes it actually livable.

Your existing account. Your existing setup. Terminal AI just gets out of the way and lets it work.

You Describe what you want. Walk away.
↓
Terminal AI Interface  ·  Skills  ·  Context triage  ·  Sessions  ·  Agent panel
↓
cc/
Claude Code Anthropic's agentic engine. The actual execution.
cx/
Codex OpenAI's coding agent. Switch providers.
Why this matters
✓
No new trust required
Claude Code is Anthropic's own tool, battle-tested and actively developed. Terminal AI inherits all of that credibility.
✓
Use your existing account
No new API keys. No new billing. Terminal AI runs through your existing Claude Code or Codex setup. Bring your own accounts. Install once, run everything.
✓
The interface Claude Code never had
Skills, Focus dashboard, live agent progress, Side Quest, persistent sessions. All the workflow Claude Code leaves up to you.
✓
Stays current automatically
When Anthropic improves Claude Code, you get it. Terminal AI rides the wave. You never fall behind the model.
✓
It gets confident before it acts
Terminal AI does its own self-analysis as it works, so it’s sure before it touches your code. Yes, it feels a beat slower than raw Claude Code. That beat is the point.
From a real session

“I think you should just use Terminal AI, dude. Just from the way you’re navigating here, it tells me that you should just do that instead.”

AH Adrian Head to a friend who was stuck — unprompted, on a call

Nobody wrote this for the website. It’s a power user routing a friend to the tool the moment the friend got stuck. The best pitch is the one you don’t make.

Features

Every tool you need.
Nothing you don't.

The best practices you'd normally spend hours configuring. Baked in. You shouldn't have to wrestle your setup just to stay organized.

📸
Screenshot In, Answer Out
Drag a screenshot in, or paste with ⌘V. Stop writing fragile prompts. Hand over the real state of the work and let the agent figure out the rest.
drag · paste · ⌘V
⚡
Parallel Agent Runs
Run multiple tasks in parallel. Watch them tackle your project from different angles. Come back to done.
live agent panel
🧠
Skill System
Teach Terminal AI your conventions and preferences once. It applies them automatically on every run. Forever.
skills/ · commands/
🎯
Focus Dashboard
Surfaces your most relevant project based on recent activity and git sync status. The right context, always at the top.
garden · focus · classic
🔍
Side Quest
A separate scratch window for quick questions that won't interrupt your main session. Ask, get an answer, keep going.
⇧⌘E
💾
Persistent Sessions
Pick up exactly where you left off. Windows, tabs, and conversations survive quit and relaunch. Your history is always there, always searchable.
window restore · searchable
⚓
Keyboard-First Flow
⌘K palette, ⌘T new tab, J/K navigation. Every action is one chord away. Your hands never leave the keyboard.
⌘K · vi nav
💡
Cross-Session Memory
Work compounds. Sessions persist across restarts, context is ready the moment you return, and coming back to a project feels like picking up a thread — not reopening an archive.
per-project · global
⚙️
Stall Detection
Your AI shouldn't stall out mid-journey. Five live states — working, waiting for you, stalled, ready, failed — so you can glance and leave again. Smart retry with backoff recovers what it can before you even notice.
5-state indicator · auto-retry
🎤
Voice Operator
Talk, hear a gist of your draft, approve with Yes. Interrupt mid-readback to tweak or reject. Your hands stay off the keyboard and the work still moves.
talk · review · approve
How it works

Set it.
Walk away.
Come back to done.

One core loop: you provide intent, agents provide execution. You don't have to watch the sausage get made.

01
Open your project
Terminal AI reads your project structure, git history, and loaded skills to build a complete picture of your codebase. Instantly.
02
Show it, don't spell it out
Drag a screenshot straight into the window, or just type. Terminal AI reads what you meant and plans the work before writing a single line of code.
03
Agents get to work
Parallel agents tackle the problem from multiple angles. Watch the agent panel, or close the laptop. Come back to done.
04
Open tomorrow like you never left
Open a session from yesterday and you're back in the room. The dev log shows the major decisions. Explore shows every step. A recap card catches you up the moment you return to a project. Ten-minute "wait, where was I?" is gone.
Agent Run  ·  feat/dashboard ● 2 active
diagnose:nav
0:42
impl:sw-cache
1:18
write:tests
0:51
deploy:build
·
Ship Log

What we shipped

Jun 20, 2026
Latest
New Powers
  • Updates install themselves — stop reinstalling by hand
  • ⌘T lands on the dashboard, not straight into chat
  • Stop chasing the scroll — toggle auto-follow mid-response
  • Find every setting in one gear menu
  • Skills and commands ready the moment you launch
More Reliable
  • Duplicate a tab and keep your actual project, every time
  • Fire your first prompt instantly, even in massive repos
  • Scroll the Explore drawer freely while agents run live
  • See what actually broke, right in the error header
  • Model chip stops squishing its cpu icon
  • Spawned sessions ride your subscription, never burn API credits
Under the Hood
  • Smoother scrolling with a wall of tabs open
  • Path errors fix themselves instead of cluttering Explore
  • No more stray input card interrupting a run mid-flow
  • Updates download and install themselves, notarized and signed
Jun 19, 2026 10 improvements →
New Powers
  • Stop auto-scroll mid-response by toggling off auto-follow
  • Skills and commands now built in, nothing to install
  • ⌘T lands on a clean dashboard, not a stray chat
  • Every setting in one gear menu, not scattered everywhere
More Reliable
  • Watch the Explore drawer stay put while agents run
  • Open large repos without the startup stall
  • Read error headers that name the real failure
Under the Hood
  • Stop being your project's scroll babysitter while juggling tabs.
  • Fewer false alarms from errors that fix themselves.
  • Killed the NudgeCard that interrupted you mid-run.
Jun 18, 2026 17 improvements →
New Powers
  • Authorize MCP servers right in the app, no terminal trip
  • A banner flags any MCP server that needs sign-in
  • The MCP auth pane confirms itself and closes once you're in
  • Jump back into any past chat from a tab
  • Long prompts scroll inside an input that fits your window
  • A pulsing beacon drops you to the latest while it streams
  • Open the app with no Gatekeeper warning, signed and notarized
  • Voice operator verifies its own work and remembers your rules
  • Answer a question card and add a typed note in one send
More Reliable
  • Cold-start a chat without the opening stutter or scroll gap
  • Recap and banner cards stop shoving the input off-screen
  • The MCP auth banner stays quiet until a server needs it
  • Worktree creation no longer fails silently from the app
  • Voice operator catches your spoken tweaks and send commands
Under the Hood
  • Only skills with a real payload show an Install button
  • Builds bundle their skill payloads reproducibly
  • MCP auth checks run in each tab's own project directory
Jun 12, 2026 6 improvements →
New Powers
  • Press ⌘N from an empty screen, a chat window opens
  • Fresh installs include problem-framer so vague ideas get shaped first
More Reliable
  • Stop losing your first message in a new chat
  • Scroll holds steady instead of yanking as responses stream
  • Scroll up mid-stream to read back; auto-follow waits for you
  • Explore panel stays clean, no empty thinking rows
Jun 11, 2026 11 improvements →
New Powers
  • Course-correct a run mid-flight with cmd-J — no stopping, no restart
  • Watch code blocks run live instead of guessing if they hung
  • Paste your key, start talking — voice ready in seconds
  • Voice keys locked in Keychain, never exposed in your shell
More Reliable
  • Stop re-picking your model every time you open a window
  • Hit Stop and the run actually stops
  • Open Settings or the voice panel without the glitches
  • Codex now finds your nvm-managed install
  • Launch straight into Opus 4.8, no setup
Under the Hood
  • Pick only the models your account can actually run
  • Find what you need without hunting through cluttered menus
Jun 07, 2026 13 improvements →
New Powers
  • Quit and relaunch without losing your windows, tabs, or chats
  • Run two models side by side in the same project
  • Hear the gist, not every word, then approve and move on
  • Cut voice off mid-readback with Yes, No, or Tweak
  • Pick Opus 4.7 or 4.8 in the model picker
More Reliable
  • Voice keeps listening after a session expires, no more silent stalls
  • Hear your draft read back the instant it's composed
  • Speak a follow-up tweak without the confirm gate swallowing it
  • Record straight to MP3 without "Encoder not found" failures
  • Find the right debug log on the first try
Under the Hood
  • Voice keeps listening after a reconnect — no more silent sessions
  • Install buttons show only skills that actually exist
  • Spot the stream-debug log path right in comments and output
Efficiency

Master the keyboard

Command palette
⌘K
New chat
⌘N
New tab (instant)
⌘T
Side Quest
⇧⌘E
Return to dashboard
⌘D
Cancel / close
ESC
Checkpoint session
⇧⌘S
Close all panels
⇧⌘W

Stop babysitting.
Start shipping.

macOS  ·  Works with your existing Claude account  ·  No new setup

Get TerminalAI $9  →

One-time purchase. macOS 14+. Requires Claude Code CLI or Codex.