thoughtdag
View on GitHubYour thinking deserves a map: an infinite canvas where LLM conversations grow into an editable thought graph. Wires are the context.
Desktop/CLI app that turns LLM conversations into an editable thought graph where wires define the model's next context. Includes a read-only MCP server to search local Claude Code, Codex, DeepSeek Harness and Pi sessions.
Use Cases
Visualize LLM conversations as an editable graphControl exactly what context the model sees nextSearch past Claude Code/Codex/DeepSeek Harness sessions by file or phraseExpose local conversation recall to agents via read-only MCP toolsRead PDFs/papers and ask questions inline with page citationsMerge, prune and branch conversation pathsExport thought-map snapshots of a research threadBenchmark how context edits change model answers
Built With
- Language
- TypeScript
- Frameworks
- React · React Flow (@xyflow/react) · Vercel AI SDK · Electron · Express · Zustand · Tailwind CSS · Vite · Vitepress · PDF.js · MCP
Tags
infinite-canvas · thought-graph · context-engineering · llm · mcp · local-first · knowledge-graph · react-flow · pkm · human-in-the-loop · dag · agentic-search · self-hosted · ollama · note-taking · session-search