clear bobcat
https://github.com/Quackstro/openclaw-brain
Still flushing out bugs 🐛!
A semantic memory layer for your AI assistant. Drop thoughts in, get organized knowledge out.
What it does:
• 🧠 Capture anything — one thought per drop, no organizing needed
• 🗂️ Auto-classification — routes to buckets (people, projects, ideas, admin, finance, health, goals, documents)
• 🔍 Semantic search — find things by meaning, not keywords
• ⚡ Action extraction — detects todos, reminders, follow-ups
• 📋 Audit trail — see how items were classified and modified
The hook:
"Tell your AI once, it remembers forever — and actually finds it when you need it."
It's the difference between an assistant that forgets everything between sessions and one that builds a working model of your world. Projects, contacts, random ideas at 2am, health notes, financial decisions — all searchable, all connected.
Why it matters:
Most "AI memory" is just chat history search. Brain is structured recall with semantic understanding. It knows that "the neighbor with the 3D printer" and "the security company guy" are the same person.