Tree

Tree is persistent, structured memory for the AI. Instead of dumping everything into one flat conversation that gets forgotten, a tree organizes knowledge into branches so the AI only pulls in what it needs for each interaction. No wasted tokens, no lost context.

It works like a real brain. Information lives in a hierarchy, gets compressed and summarized over time, and the AI navigates to the relevant parts when you talk to it. Both you and the AI can read from and write to the tree, so the context evolves as work progresses.

You can have as many trees as you want, each one for a different project, topic, or area of your life. Invite collaborators to build together, use it solo to stay organized, or design systems that others can follow. Tree works however you need it to.

This is a personal project by Tabor Holly.

It's my attempt at building what I believe large language model technology will need going forward: persistent memory, a navigable and fluid GUI driven by text input, and a unified structure where instructions, knowledge, and history all live together — readable by both humans and LLMs.

Dive deeper into how Tree works and what you can build with it.