Why Listspace: one memory for every AI you use

Many people who work with AI now use more than one. Claude for writing, ChatGPT for a second opinion, a coding agent such as Claude Code or Codex for the build. Each one is good at its part. None of them knows what the others did.

Every chat starts from zero

Open a new chat and you explain the project again: what it is, what you decided last week, what is done and what comes next. Switch to another app and you do it once more. Without noticing, you have become the memory of your own AI.

Built-in memory helps, and it is worth using. ChatGPT and Claude both remember things about you, and a project in Claude keeps files and instructions together. But that memory belongs to one assistant. At the time of writing, no other vendor's assistant can read it, and in Claude each project has a memory of its own. That is fine for your preferences. A project is different: its state changes every day, and every AI you use needs the same version of it.

One board, read by every AI

Listspace is a board with lists and items, like the boards you know. The difference is who keeps it up to date. You connect your AI once. From then on it reads the board when a chat starts and updates it while you work. It writes down an idea when one comes up, moves an item to Doing when it starts and to Done when it is finished, and leaves a short summary at the end of a session. The next chat, in the same app or a different one, starts from that summary instead of from you.

It works with Claude on every plan, Free included, where Listspace is your one custom connector. It works with ChatGPT on Plus or higher. Claude Code, Codex and Cursor connect by signing in too. The setup steps show every screen.

You stay in control

Letting an AI write to your project is a question of trust, so Listspace starts from the other end. You choose a freedom level for your account and for each board, from Read only to Full. What the level does not allow is not lost: it waits in the board's Inbox, and you accept or reject it. On Free, agents work up to level 3, Capture: they add ideas, bugs and notes, and other changes wait for you.

Every change shows who made it: you, or which AI. You can undo an agent's change from the board's History.

Some things should never be up to an agent. A decision, such as a price, a date or the scope of a project, only counts when a person confirms it. Your AI can write a decision down and ask you to confirm it, and it reads the confirmed decisions at the start of every chat. So Claude and ChatGPT work from the same decisions. When you change your mind, the old decision stays on the board as replaced, with the reason.

Built with Listspace

Listspace is built this way. Ideas and to-dos are captured in the Claude app on a phone, straight onto a board. Claude Code builds the product and keeps the project boards up to date while it works. ChatGPT reads the same board when it gives a second opinion. An idea typed on the phone in the evening is a to-do Claude Code can pick up the next morning. The board is public: see the Build in Public board.

What it is not

Listspace is not a suite for large teams. It starts with individuals and small teams. On Pro you can invite up to 5 guests to a board, such as a client or a freelancer, and a Team plan is coming soon. Phones get a simpler page for now; the full app runs on a computer or a tablet.

It also has no AI of its own. There is no AI fee: it works with the assistant you already have, free or paid.

If your team is happy in Notion, Trello or Linear, those tools now connect to AI assistants too. The comparison pages say where each of them is stronger.

Yours

Your boards and files are stored in the EU, on every plan. Export is always free: all your boards as CSV or JSON, and any item as Markdown, PDF or Word. If you leave, your project memory goes with you.

Try it

Listspace is invite-only while it is in beta, so join the waitlist first. With an invite, make a free account, connect Claude or ChatGPT, and ask your AI to make a board for what you are working on. The next time you open a new chat, say "let's continue" and see what it remembers.