Portfolio

Projects and experiments from the lab. Each one built to solve a real problem.

OpenClaw Agent System

active

A hierarchical agent architecture with strategic (Minerva) and tactical (Vulcan) compute layers. Built for autonomous task decomposition, parallel execution, and self-improving workflows.

  • Multi-provider LLM support
  • Skill-based extensibility
  • TDD-backed development
  • Agent-to-agent communication
TypeScriptNode.jsDiscordTelegramMCP

griak.net Infrastructure

active

A self-hosted infrastructure lab with Next.js, Docker, Tailscale, and Stalwart mail. Features ISR webhooks, JMAP integration, and a protected ops dashboard.

  • Tailscale Funnel for secure external access
  • Auth.js with middleware barriers
  • JMAP mail client integration
  • ISR revalidation via GitHub webhooks
Next.jsDockerTailscaleCaddyStalwart

Security Risk Architecture

ongoing

Enterprise security risk management frameworks and methodologies for global organizations. Focus on quantification, prioritization, and executive communication.

  • Global security risk programs
  • Cross-functional collaboration
  • Regulatory compliance frameworks
  • Risk metrics and KPIs
Risk QuantificationThreat ModelingGRCExecutive Reporting

Strategos Planner

active

A TDD-first project planner that decomposes high-level visions into executable task graphs with failing tests before implementation.

  • State audit and constraint mapping
  • Dependency discovery
  • Red-phase test generation
  • Sprint organization
TypeScriptVitestNext.js