ShaneBrain Learning

Auto-learning inbox pipeline: drop a file in, it gets ingested to Weaviate on cron. Zero-friction knowledge capture for the ShaneBrain brain.

View on GitHub Treasures Hub
Python Weaviate Ollama Cron Pipeline

How It Works

Drop file in inbox/
→
Cron picks it up
→
Embeds with Ollama
→
Stores in Weaviate
→
Moves to processed/

Features

Zero-Friction Ingestion

Drop any document into the inbox folder. The cron pipeline handles everything automatically — no scripts to run, no commands to type.

Weaviate Integration

Uses nomic-embed-text via Ollama for 768-dim embeddings. All knowledge lands in the LegacyKnowledge collection.

Audit Trail

All ingest runs logged to logs/. Processed files moved to processed/ to prevent double-ingestion.

ShaneBrain Ready

Works alongside the ShaneBrain MCP server. Ingested knowledge is immediately searchable via the 42-tool MCP interface.

Folder Structure

shanebrain-learning/
├── inbox/        # Drop files here — auto-ingested
├── processed/    # After successful ingest
└── logs/         # Ingest run logs

ShaneBrain Ecosystem

AI-Trainer-MAX

5-phase local AI training curriculum for non-technical users

angelcloud-actual

Angel Cloud platform — Firebase, Pulsar security, logistics API

thought-tree

React visual thought browser with Weaviate + Ollama semantic search

srm-dispatch

SRM Concrete North Alabama daily dispatch tool — React PWA

mini-shanebrain

Multi-platform social bot framework (X, FB, LinkedIn, etc.)

multi-container-app

Docker multi-container template (Node.js + MongoDB)

treasures

ShaneBrain recovered artifacts hub — scripts, projects, creative
Built with Claude + ShaneBrain — AI tools for humans who build.