Family Brain — Live Network Visualization

Status: ✅ Implemented & live at https://hoffdesk.com/api/system/brain/svg
Preview: https://hoffdesk.com/blog/static/brain-live.html
Stats JSON: https://hoffdesk.com/api/system/brain/stats

Overview

A live-rendered SVG network visualization of the Beelink's data constellation. Each node represents a data store, auto-scaling with live record counts. Connections pulse with animated data packets. No manual SVG editing — the server generates it from real queries.

Nodes

Node	Data Source	Auto-scales With
BEELINK	—	Static hub (r=58)
CALENDAR	Radicale systemctl status	Status text
MEMORY	ChromaDB chunk count + file size	Chunk count → radius 36-48
BLOG	SQLite post count	Published + total → radius 36-48
LLM	Ollama API model count	Model count → radius 36-48
ONTOLOGY	Entity count	Entity count → radius 34-42
RECIPES	Ontology recipe-type entities	Stub (0 now, grows as recipes added)

Endpoints

GET /api/system/brain/svg → image/svg+xml

Renders full network SVG with live data. No auth required. Used on dashboard as <img> or <object>.

GET /api/system/brain/stats → JSON

Returns raw stats for each data store. Useful for debugging or embedding in other dashboards.

How to Add a New Node

1. Add a query to _get_db_stats() in /home/hoffmann_admin/.openclaw/workspace-socrates/hoffdesk-api/family/brain.py
2. Add a color entry in COLORS dict (use one of the 7 existing palette colors)
3. Add a slot in _build_nodes() with position and detail text
4. Restart uvicorn — the SVG auto-generates

Deployment

The endpoint is mounted on:
- family_app (family.hoffdesk.com) — with session auth exempt prefix
- blog_app (hoffdesk.com) — fully public
- dev_app (local testing) — fully public

Auth exemption added to EXEMPT_PREFIXES in /api/system/.

Files

File	Purpose
family/brain.py	Stats collection + SVG generator
blog/templates/brain_svg.svg.j2	Jinja2 SVG template
hoffdesk/blog/static/brain-live.html	Preview page with HTMX auto-refresh
shared/project-docs/family-brain-spec.md	This spec

Phase 2 Ideas (Socrates)

• Add query_count tracking to each service (read from FastAPI access logs or middleware)
• Add last_activity timestamps for stores that don't track them
• Add health check per store (ping port, check process)
• Add cross-connections between satellites (e.g. Blog → Memory for vector search)