Your AI agent works on your Mac until it reaches a step only you can do, like an Allow prompt, a 2FA code or a payment. Then it prints "click Allow" into a terminal you aren't watching. bigarrow gives it a finger: it draws a big arrow with a sign on top of any window, lets your clicks pass through and removes itself.
Who this is NOT for
It won't help if your team works on Windows or Linux, since it is macOS only. Same if you don't yet run agents like Claude Code or Codex from the terminal: without an agent giving the orders, the arrow has nobody to obey. And if you want a bot that clicks for you, this isn't it: the README says it never clicks, types or captures the screen. It only points.
What matters is what it chose not to do. An agent that can press Allow for you is a risk; one that points at the button and waits for your decision isn't. Permissions, payments and signatures stay in your hands. For a small business the best payoff isn't the agent, it's the manuals. With `--png` you can build internal guides from marked-up screenshots ("this is how you export the invoice to PDF") without fighting an image editor. Heads up: the repo was created on October 8, 2026 and is two days old. It has CI and zero open issues, but it hasn't been through long-term use yet.
Your first step
On a Mac with Homebrew, run brew install franzenzenhofer/tap/bigarrow, then try bigarrow point --at 760,500 --text "Here" to see the arrow for 8 seconds. If you like it, bigarrow install-skill teaches it to Claude Code.
The big-arrow-on-the-screen (bigarrow) implementation guide
The project belongs to Franz Enzenhofer (franzenzenhofer) and it is free: the link beside asks for nothing. The guide to set it up without getting lost is ours: step by step, common mistakes and a checklist. Leave your email and we will send it.
No spam. Unsubscribe anytime.