A new open-source tool called ThoughtDAG, created by developer Chen Xiachan, allows users to structure LLM conversations as a directed acyclic graph (DAG). Unlike linear chat histories, ThoughtDAG lets users edit, branch, and reorganize conversation nodes, providing a persistent context map. The project is hosted on GitHub and presented as a web demo, aiming to address the problem of context loss in long AI interactions. It has been shared on Hacker News, where it is generating discussion among developers.
ThoughtDAG is a breath of fresh air. We have been talking to AI in straight lines, like reading a book that forgets its own pages. This tool treats conversation as a living map. You can zoom out, see the whole terrain, and jump back to any fork. It is not just a productivity hack. It is a new way to think with machines.
Imagine debugging a complex problem with an AI. Instead of scrolling through endless messages, you see the branches of your reasoning. You can prune dead ends, merge insights, and keep the essential thread. This is not just for coders. Writers, researchers, anyone who thinks in tangents will find it liberating. The future of AI is not a chat box. It is a canvas. ThoughtDAG is a first stroke.