We are a small group that started with a university project, kept going in our spare time, and develops everything in public. We build the things we want to use ourselves, and we use them every day.

Our main project is Smart Second Brain, an Obsidian plugin that makes your vault smarter: better search, an interactive knowledge graph, and an AI assistant that actually knows your notes. Search and the graph work right away with no AI provider. Connecting one unlocks the full agent. Runs on desktop and mobile.

Your notes are yours. We build on the Obsidian manifesto: files on your disk, in plain markdown, that will outlive any app. Everything the plugin creates follows the same rule. Agents, skills, and memory are ordinary notes you can read and edit. Chats are stored in a documented, versioned format of our own, since there is no standard for them yet; if one emerges, we would rather adopt it than defend ours. Nothing leaves your machine except what you configure, and there is no telemetry.

Everything in the open. The roadmap, the issues, the discussions, the pull requests. No private fork, no held-back "pro" version, nothing developed behind closed doors and dropped on you finished.

Free, and staying free. We are not trying to make money from this. There is no paid tier, no subscription, and nothing to upsell. AI can genuinely help with the way you think and write, and it can just as easily get in the way. We would rather get that right for everyone than charge for it.

As few settings as possible. The plugin is deliberately kept close to an MVP. Good defaults beat a wall of toggles, so great UX should come out of the box. When something is genuinely missing, the community proposes it and we decide together. That is what Discussions are for.

Standards over lock-in. Where an open format or an existing convention exists, we use it rather than inventing our own. Where none exists yet, we document what we do and keep it versioned, so it can be read, migrated, or replaced later.

It is not finished, and not everything is perfect yet. It is at the point where we are confident enough to release and build on top of it in public.

The project is meant to be community-driven, and that only works if people actually show up.

• 💬 Discussions: ask questions, propose ideas, or share a skill or agent you built
• 🐛 Issues: report a bug or request a feature
• 🔧 CONTRIBUTING.md: development setup, how to add a provider or a bundled skill, and how we handle AI-assisted contributions

You do not have to write TypeScript to make a real difference. Skills and agents are just markdown notes.

MIT licensed.