hrdx: I Built My Own Agent Multiplexer, Because of Course I Did
On building hrdx, a minimal terminal multiplexer for coding agents in Go. Not to compete with anything, but because I wanted my own tool, again.
Writing by Patric Eckhart.
Live Last read · last published · next check
On building hrdx, a minimal terminal multiplexer for coding agents in Go. Not to compete with anything, but because I wanted my own tool, again.
A portable, permission-scoped packaging format for AI agents. Persona, skills, requirements, and enforced permissions in one .zot artifact that runs locally with your own keys.
How zot ships its internals as importable Go packages, so you can write a working coding agent harness in about a hundred lines.
A counterintuitive case for the cheap plan. The usage cap on ChatGPT Plus need not be a downgrade, it can be a speed limiter, and a speed limiter can be exactly what most of us need right now. Myself included.
A small browser automation CLI for coding agents, built as a learning vehicle for understanding how agents should see and use the web.
A small, private, on-device voice-to-text app for macOS. Hold a hotkey, speak, release, paste. Here is why I built it and how it turned out.
On building my own coding agent harness from scratch, what I wanted to learn from it, and why Go turned out to be the right choice.
A concise, no-nonsense guide to understanding Event Sourcing from first principles. Neutral, practical, and open.
A compact, type-safe TypeScript framework for CQRS and Event Sourcing with pluggable database adapters and zero runtime dependencies.
Thoughts on a declarative, file-based access control model for GenesisDB, and what a single thoughtful Hacker News reply taught me about it.
GenesisDB is a lightweight, production-ready event-sourcing database engine written in Go. Here is what it does, what it doesn't, and why I built it.
Event streams are append-only by design. GDPR demands deletion. Here is how I think about reconciling the two, and the approach I prefer in practice.
Thinking through what it would take to build my own event store: storage layout, performance, indexing, and the part that scares me most, write locking.