I build practical security, monitoring, and automation tools for homelabs and small teams. Technical operations and customer success background, hands-on with networks, servers, endpoints, and AI-driven workflows. Everything below is open source, tested, and actively maintained.
| Project | What it does | Stack | Status |
|---|---|---|---|
| eagleeye2 | OSINT: find a person's social profiles from a face photo | ||
| endpointwatch | Live-response Windows endpoint monitor with LLM alerts and baseline drift detection | ||
| pinggrid | Network monitoring dashboard that auto-configures itself | ||
| rackpulse | Health monitoring for legacy Dell PowerEdge servers on ESXi | ||
| dellsilencer | PowerEdge fan control with hysteresis, no more oscillation | ||
| restore-drill | Verify your backups actually restore, in one command |
| Project | What it does |
|---|---|
| phonepilot | ESP32 AI agent platform for iPhones: autorun workflow templates, a social posting queue with confirmation, and a fleet manager |
| agentpulse | Operations dashboard for agent fleets: cron health, drift, errors, one dark page |
| esp32-mcp | MCP server that gives AI agents GPIO, relay, and sensor control of ESP32 hardware |
| agent-surface-report | Windows audit: what an AI agent could touch and how it could persist |
| re-ai | Reverse engineering as a structured writeup: PE parsing, entropy, LLM analysis |
| mcp-firewall | Policy gateway and audit log for MCP tool calls |
| diagrams-from-json | Architecture diagrams from a spec file: SVG, layers, labeled edges |
| trend-radar | What is trending on GitHub and HN, scored against a YAML skill profile |
| ollama-watch | Local LLM monitoring: models, throughput, daily digest, dark dashboard |
- The October 2026 fleet: release cadence, docs, and real hardware runs.
- Security: TryHackMe and HackTheBox labs, OSCP in progress.
- Open to senior technical operations, customer success, and field engineering roles.


