Thinking Crutch · Free · Runs on your Mac
Just talk with your agent. As you go, it puts the questions, ideas and reasons that matter on a canvas, so you can see how you got here. What counts is your call.
Thinking Crutch keeps only what is worth keeping, and every decision can be traced back to why.
1 / 6
Did that hit home?
An Idea keeps its reasons; a Conclusion keeps its basis. Only what is still unresolved moves on to the next round.
What's worth keeping is already on the canvas. Next time, just start a new conversation and let your agent read it first. The old chat? Throw it away.
Free. Your canvases stay on your computer.
Signed and notarized by Apple, so it opens with a double-click.
Download for MacmacOS 12 or later, Apple Silicon and Intel.
A Windows version is on its way.
Coming soonAbout five minutes from download to your first canvas.
Download Thinking Crutch for Mac and open it. Keep it open while you work: your agent reads and writes the canvas through it.
Send this to your agent. It adds the canvas to its own MCP configuration.
Please add this MCP server to your own configuration: http://127.0.0.1:47821/mcp (HTTP transport, name: ai-canvas). When you're done, remind me to start a new session so you can use it.
Or add it yourself. The address is http://127.0.0.1:47821/mcp
Tell your agent what you want to think through. It writes the goal, questions and ideas onto the canvas as you talk.
“Read my canvas and help me think about this goal.”
Everything your agent writes waits on the canvas until you Apply it. Reject what doesn't belong.
What people ask before they try it.
On your computer, as files. There is no account: your canvases and everything you work on stay on this machine.
No. It's free.
Any agent that can connect to an MCP server over HTTP. We use Claude Code and Codex every day.
No. It proposes; nothing counts until you Apply it.
The agent endpoint listens only on this machine (127.0.0.1). It has no sign-in yet, so other programs on your computer can reach it too.
Not yet. Mac comes first.
Write to us. Where you got stuck is the most useful thing to tell us.
What helped, what felt confusing, or what was missing? Your feedback shapes what we improve next.