Projects Overview

Here are my projects — from websites to infrastructure and developer tooling.

Working Projects

Clawdie ISO — Operator USB Image (Main Project)

Bootable USB image — development surface for Colibri + installer for colibri service on bare metal. This is the most significant project, combining FreeBSD operating system, ZFS, bhyve virtualization, Bastille jails, and Colibri control plane into a single portable artifact.

Technology Stack: FreeBSD 15 (quindecim) · XFCE · ZFS RAID1 mirror · bhyve · Bastille jails · pgvector · poudriere · age-encrypted seed partition

Key Milestones: View full build log in wiki →

Seed partition + age encryption (PIN unlock, no Vaultwarden)
Firstboot pipeline: seed decrypt → Pi agent (npm global) → home overlay
Colibri-daemon (systemd-style, AUTOSPAWN=YES, PostgreSQL + pgvector)
Mother/Hive architecture: colibri-mcp SSH-stdio bridge + tool proxy
Poudriere package repo + verified artifacts (build-manifest.json, check_release_gate)
Hardware-validated on Valhala (HP ML350p Gen8)

Current Version: 0.12.0 (released) → 0.13.0 target (in development: Colibri integration, wiki, OKF)

Build Host: Valhala — HP ML350p Gen8, FreeBSD 15, doas, poudriere bulk builds

Colibri — Tokenomics Trifecta (Rust + Go) — Work in Progress

Control plane for autonomous AI agents with tokenomics trifecta: performance / speed / cost. Target metric: cost-per-intelligence > cost-per-token.

⚠️ Note: Still in active development. Components work individually, but full chain (boot → seed → Pi agent → Colibri daemon → Mother/Hive → benchmark) not end-to-end validated on hardware.

Technologies: Rust (core: daemon, CLI, zot) · Go (services: API, MCP bridge, cost dashboard) · PostgreSQL + pgvector · bhyve VMs · Bastille jails

Components: View status matrix in wiki →

colibri-daemon (Rust) — core loop, systemd-style supervision ✅
colibri-cli (Rust) — basics, missing advanced subcommands 🔄
zot agent harness (Rust) — NDJSON RPC, subtask delegation, tool interception ✅
colibri-mcp (Go) — SSH-stdio bridge, tool proxy ✅
cost-dashboard (Go) — in development 🔄
llama.cpp benchmarks (Rust/Go) — /v1 timings object for ms/tok ✅

Documentation: Colibri wiki — 44 pages (LLM benchmarks, architecture decisions, OKF v0.2, pre-push hooks)

Blockers for end-to-end: bhyve spawn logic in ISO 🔄 · pgvector migration script 🔄 · cost-dashboard integration 🔄 · hardware test on Valhala 📅

Herdr — Terminal Workspace Manager

Technologies: Rust, ratatui, PTY, tmux-style sessions

Description: Terminal-native agent runtime and multiplexer. Runs separate AI agents in real tabs, enables session attach/detach, uses SSH attach, agent state (blocked/working/done), and an API for orchestration.

Status: Active development, published at herdr.dev · GitHub

Zot Agent Harness

Technologies: Rust, NDJSON RPC, Codex/Claude/OpenCode integration

Description: Agent harness for delegating coding tasks. Enables zot rpc communication, subtask delegation, tool-call interception, and multi-agent workflows.

Role: Foundation for Hermes-BSD multi-agent orchestration

Pi Harness

Technologies: Node.js, @earendil-works/pi-coding-agent (npm global), Aider

Description: Standalone, configurable agent for autonomous development. Primary execution lane on Clawdie ISO — installed as npm global with home overlay (.pi/agent/ → $HOME). Runs under Hermes orchestration (Hermes=orchestrator, Codex=FreeBSD/ISO, Claude=design, Zot=coding). Replaced OpenClaw (Peter Steinberger) as the suitable harness for Hermes ecosystem.

Websites

Personal Website (this site) — Astro, Starlight, TypeScript · Multilingual CV/portfolio · In development

Bučko - Dog Website — Astro · bucko.smilepowered.org · In production

Infrastructure

VPS Server on Vultr — Docker, Traefik, Linux · Self-hosted web applications with reverse proxy and SSL certificates · Services: websites, n8n (workflow automation)

Currently Learning / Research (merged)

Actively learning and exploring:

Rust — systems programming, Colibri core, Herdr, Zot
Go — Colibri services, CLI tools, high-throughput services
Python — for use in AI/ML projects
LangChain / LangGraph — tools for building RAG systems
n8n — workflow automation
Local LLM models — experimenting with Ollama
MCP Protocol — Model Context Protocol for tool integration
Blender Geometry Nodes — procedural modeling of geodesic domes
FreeBSD 15 / ZFS / bhyve — advanced disk management, jails, virtualization
Hermes Agent — personal AI agent across CLI, TUI, gateway, desktop
Claude Code / Opencode / Codex — AI coding assistants
RAG Systems / pgvector — vector memory
Tokenomics (perf/speed/cost) — cost-per-intelligence metric
llama.cpp timings — ms/tok measurement
Google OKF — Open Knowledge Format frontmatter

Updated: 2026 | Projects and interests continuously evolving.