Thoughts captured. Focus unbroken.
A developer-first, outliner-based note-taking tool that stays out of your way
You're deep in your work. An idea surfaces.
Most note apps make you stop what you're doing: open the app, find the right place, format it properly. By the time you're done, your train of thought is derailed.
Flow is different.
flow capture "Research CRDT sync options for offline-first #idea"
# Done. Back to work.Your thought is captured. Your momentum continues. Find it later when you need it.
- Local-First — Your data stays on your machine, in plain Markdown
- Outliner-First — Think in bullets, expand to prose when needed
- Developer-Friendly — CLI-native, scriptable, fits your terminal workflow
- No Lock-In — Your notes are just files, take them anywhere
| Component | Description |
|---|---|
| CLI | Quick capture and search from your terminal |
| TUI | Full-featured terminal interface |
| GUI | Native desktop app for longer sessions |
| Server | Self-hosted sync across devices |
Mix and match — use only what you need.
- Project X #project
status:: 🟩 Active
- Architecture decisions
- Use microservices for scalability
- [[PostgreSQL]] for main DB
- Tasks
- [ ] Set up CI/CD pipeline #task
- [ ] Write API documentation #taskEvery bullet is a node. Add tags to create objects you can query and connect across your knowledge graph.
⚠️ Flow is in early development. Star and watch to follow progress!
- Project structure and architecture
- Core storage layer (Markdown + Loro CRDT)
- CLI integration
- TUI interface
- GUI application
- Sync server
# Coming soon — Flow is in active development
# Star the repo to get notified when it's ready!📖 Read the documentation for more details.
Contributions are welcome! Please read the Contributing Guide to get started.
Flow draws inspiration from Logseq, Tana, Obsidian, Roam Research, and jrnl — combining the best of outliners, knowledge graphs, and CLI simplicity.
This project uses AI as a development tool — all AI-generated content is reviewed and refined by a human.
I understand that AI in software development is a divisive topic. Many developers feel it undermines code craftsmanship and the art of building software by hand. I respect that perspective. However, as AI tools become standard across the industry, choosing not to use them means falling behind. This is my first project where I'm utilizing an AI coding agent, and I'm approaching it as a learning experience.
AI assists with:
- Exploring ideas and concept refinement
- Generating boilerplate code
- Drafting documentation
AI does not write:
- Core logic and algorithms
- Architectural decisions
- Critical code paths
I believe in transparency about AI usage while maintaining quality standards.
⭐ Star this repo if you're excited about Flow!