Tools
Open source infrastructure for the research projects.
"I built an external brain because context switching was destroying me."
Core Infrastructure
The core cognitive infrastructure. 359K messages queryable in 256ms. Semantic search across 106K embedded messages. The external brain that makes everything else possible.
For anyone building personal knowledge infrastructure or AI memory systems.
Local-first video transcription. Whisper/Parakeet ML. Zero API costs at scale. Watch and forget → transcribe, embed, retrieve.
For heavy video consumers who want searchable knowledge from their watch history.
Frameworks & Tools
Mental health dimensional framework. Alternative to DSM categories. Body-based dimensions instead of symptom checklists.
For clinicians, researchers, or anyone exploring alternatives to categorical diagnosis.
Human sovereignty framework for AI systems. SHELET protocol. 100% human agency, AI in the loop.
For developers building AI systems that preserve human decision authority.
Give any LLM its own display. Send JSON, get interactive UI. Visual interface for AI outputs.
For developers who want LLMs to render rich, interactive interfaces.
MCP server for querying the Sefaria Jewish text library. Access Torah texts through AI.
For AI applications that need access to Jewish texts and commentaries.
The Brain MCP Architecture
359K raw messages (2023-2026)
↓
Embedding pipeline
↓
106K vectors in LanceDB (32x smaller than DuckDB)
↓
Semantic search (256ms)
↓
MCP server interface
↓
Claude Code / Any LLMExternal memory for minds that can't hold it all. Context preservation across sessions. The cognitive prosthetic that makes poly possible.
What The Brain Tracks
Conversations
359K messages
YouTube
32K videos tracked
GitHub
Repos + commits
Principles
8 SEED principles
Embeddings
106K vectors
Timeline
2023-2026