Skills / evidence over percentages

What I use to close the loop.

No arbitrary progress bars. Each group describes the kind of behavior I can build and points toward observable work.

02 / Frontend

React & TypeScript

Type-safe browser interfaces that render streaming state and stay decoupled from backend domain models.

  • React
  • TypeScript
  • Vite
  • SSE
Evidence: Runtime Inspector ↗
03 / Infrastructure

Kubernetes & container deployment

Deploying and operating services on a self-hosted single-node cluster (minikube) — Deployments, Secrets, hostPath storage, container images, and Cloudflare tunnels.

  • Kubernetes
  • minikube
  • Docker
  • Cloudflare Tunnel
Evidence: SearchX ↗
04 / Systems

Rust & Tokio

Typed event-driven systems — domain events, pure decision functions, reducers, actors, Tokio async, and effect execution.

  • Rust
  • Tokio
  • Async
  • Actors
Evidence: Looper ↗
05 / Platform

ArchLinux — 4 years of daily use

Four years of ArchLinux as a daily desktop and server platform — KDE and Hyprland (end-4 dotfiles), rolling-release maintenance, systemd services, packaging, and dotfile-driven configuration.

  • ArchLinux
  • KDE
  • Hyprland
  • systemd
  • dotfiles
Evidence: dots-hyprland (end-4 reference) ↗
06 / Research

Agent runtime research

Ongoing architecture research on agent runtimes — extension models, out-of-process plugins, trace-first observability, and the harness/product boundary.

  • Agent runtime
  • Plugins
  • Trace-first
  • Design
Read the design principles ↗