AMA: We rebuilt Kuku as an open-source, local-first second brain for AI

Hey everyone 👋

We’re launching Kuku today on Product Hunt, and I’d love to do a small AMA / feedback thread here.

Kuku started as an Obsidian-like Markdown editor, but we rebuilt it into something more specific:

an open-source, local-first second brain that turns your Markdown vault into reusable context for AI tools.

Why we’re building it:

AI chats are useful, but memory is fragmented.

Your useful context gets stuck inside tools, chats, or closed note apps.

Kuku keeps the source of truth as plain Markdown files, then adds:

• wikilinks, backlinks, graph, and search

• AI-assisted edits with reviewable diffs

• local-first storage

• open-source architecture

• a path toward portable AI memory across tools and models

The question we’re exploring:

What should a “second brain for every AI” actually do?

Would love your thoughts on any of these:

1. Would you trust an AI to edit your notes if every change came as a reviewable diff?

2. Should AI memory live inside chat apps, or in your own local files?

3. If you use Obsidian, Logseq, Roam, Notion, or Cursor, what would make you try Kuku?

4. What plugin or workflow would you want first?

5. What’s the biggest risk with AI-assisted personal knowledge management?

Happy to answer anything about:

• local-first architecture

• Markdown as source of truth

• AI edits

• open source

• sync

• privacy

• why we rebuilt the product

Launch page is live today — but mostly, we’re here to learn from people who care about knowledge tools and AI context.