Run BoggersTheAI
on Your Machine
Local-first. No cloud. No API keys for the core. One Python install and your laptop becomes a running TS instance.
Install in 5 Steps
~5 minutes from zero to running wave cycle.
$ git clone https://github.com/BoggersTheFish/BoggersTheAI$ cd BoggersTheAI
$ pip install -e ".[llm,adapters]"# Optional extras: faster-whisper, piper-tts (multimodal)
# config.yaml ships in the repo — edit inference.ollama.model / paths as needed
# https://ollama.com$ ollama pull llama3.2$ ollama pull nomic-embed-text
dashboard-start# FastAPI dashboard → http://localhost:8000# Or: boggers (CLI + background wave)
Dashboard Features
FastAPI + Cytoscape.js + Chart.js + Rich TUI — full observability stack.
Try It Here
A browser-based preview of the wave cycle. Click nodes, push activation, watch propagation.
Click nodes to push activation
Ask the graph
Your prompt is decomposed into concepts, matched against the live graph, and optionally explored via a probe node and wave cycles when context is thin; the LLM then synthesizes an answer. Requires the Docker backend and Ollama.
Live server graph
Same graph the AI uses: activation size, edges, wave cycle + tension. Probes from graph exploration appear in orange; conversation nodes in purple. Refreshes every few seconds and after activity on TS Chat.