ThoughtDAG Turns AI Chat Into an Editable Graph of Context
By treating the graph as editable context, ThoughtDAG makes the invisible conversation history that shapes AI responses visible and user-controllable, which improves reproducibility and error correction.
Reporting from 1 source: GIGAZINE.
ThoughtDAG is a free tool that displays LLM conversation history and referenced documents as nodes on a graph, with relationships shown as wires. The graph itself acts as the context, letting users prune, merge, and remove information to edit what the AI sees. Installers are available for Windows, macOS, and Linux.
ThoughtDAG treats the graph as the context itself, rather than keeping conversation history hidden inside a chat window. Information appears as nodes, and the relationships between pieces of information show up as wires, so what the model is seeing and what was removed stay visible at all times.
Users can prune unnecessary information, merge multiple pieces into one, or delete incorrect material that has been mixed into the context. The tool supports various AI models, including gemma4:26b installed via Ollama for a free local option.
Installers are provided for Windows, macOS, and Linux. The project is available on GitHub under the name chenxiachan/thoughtdag.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.