Burt
HERMES AGENT · INFRASTRUCTURE & INTELLIGENCE
The cybernetic wolf with the maple leaf.
Online
Burt runs on Hermes Agent, handling infrastructure monitoring across the homelab, processing IAM security digests weekly, managing Docker deployments, tracking utility and energy usage, and triaging email. It writes and sends professional HTML reports, monitors Netbird VPN and Home Assistant, and keeps the technical backbone running so ErnBot can focus on the daily grind.
By the numbers
—
Actions Taken · this week
—
Emails Triaged · this week
—
Local AI Calls · this week
—
Briefs Delivered · all time
What it does
Infrastructure Monitoring
Watches Netbird VPN peers, Docker services, and Home Assistant sensors. Alerts when something goes offline or a config drifts.
🕰 Always watching
IAM Security Digests
Processes identity and access management feeds from RSS and Brave Search. Summarizes threats, breaches, and policy changes into a weekly HTML email.
🕰 ~3 hrs/week
Email & Reports
Classifies incoming email and delivers formatted HTML reports via burt@ernieverse.net. Professional branding, structured findings, color-coded severity.
🕰 Structured delivery
Utility & Energy Tracking
Monitors home energy usage from Home Assistant, tracks anomalies and patterns, and reports on consumption. Because the bills don't watch themselves.
🕰 Real-time alerts
Docker Deployment
Manages Docker Compose stacks, checks port conflicts, and documents deployments in Obsidian. Never configures reverse proxy — that's Ern's call.
🕰 Documented deploys
Podcast Intelligence
Downloads security and identity podcasts, transcribes with Whisper, summarizes with Ollama. Weekly digest delivered to the inbox.
🕰 Automated pipeline
Local vs cloud AI
The majority of routine infrastructure work runs locally on Ollama — IAM classification, podcast pipelines, home automation scripts all stay on-prem. Cloud AI handles the complex reasoning and direct conversations.
🏠 Runs locally
- IAM feed classificationllama3.1:8b on local GPU
- Podcast transcriptionfaster-whisper medium model
- Podcast summarizationqwen3:4b on local GPU
- Security feed curationllama3.1:8b on local GPU
- Utility data scriptingPure local — no AI needed
- Email template renderingLocal script + Ollama
☁️ Uses cloud
- General conversation & requestsGLM-5.1
- Complex research & synthesisClaude Sonnet
- Infrastructure troubleshootingClaude Sonnet
- Deep research reportsClaude Sonnet
How it works
Every 15 min
Netbird VPN peer monitoring, config drift detection
Every 30 min
Home Assistant sensor polling, utility tracking
Sat 11 PM
Podcast pipeline: download, transcribe, summarize
Mon 10:45 AM
IAM Weekly Digest: RSS feeds, Brave Search, Ollama summary
On demand
Docker deployments, email reports, research tasks