I put every coding agent session into ClickHouse
How I send every Claude Code, Codex and OpenCode session from all my machines into ClickHouse, what that's good for, and a prompt to build your own.
Notes on AI systems, developer tools, and the craft of building software.
I'm Daniel Dewhurst, a Lead AI Solutions Engineer. I build with AI, Laravel, and TypeScript for a living. This is where I write about what the tools change and where they fall short.
latest
recent
My coding agents live on a headless VM, and my phone is the remote
How I run T3 Code on a Proxmox VM with no screen, reach it over Tailscale from a MacBook and an iPhone, and treat the thread list as an inbox.
I made my coding agents talk like an aircraft maintenance manual
Two weeks of agent replies in Simplified Technical English: shorter replies, list labels I can answer with, and the rules that still slip.
Migrating to Claude Opus 5.5: a cheat sheet for the API and Claude Code
Effort defaults, the four breaking changes, silent progress updates, and the prompt and CLAUDE.md lines worth changing when you move from Opus 5 to 5.5.
Story Skills: a continuity compiler for AI-written fiction
My open-source Agent Skills bundle gives coding agents a shared story bible format — and a deterministic CLI that catches dead characters walking, payoffs before setups, and unfired Chekhov guns.
Muse vs Grok Bot vs OpenClaw & Hermes: a practical personal-agent comparison
Meta's Muse, xAI's Grok Bot, and the open-source route (OpenClaw and Hermes Agent) are all fighting to be your always-on AI assistant. They disagree on who does the setup, whose model you get, and where your data lives.