A working collection of tools and ideas.Build → test → write
Tools for agents.
Notes for humans.
I build AI systems and write about what I learn. Experiments, useful tools, and the details that make them work.
Inside WireAt Steel
wire run --objective "Go to example.com and return the heading text"
wire craft --run-id <run-id> --standalone --out replay.mjsSelected writing
Browse the archive ↗-
Claude Code with GLM-5.3: The Setup That Works
Run GLM-5.3 through Z.ai on Claude Code 2.1.266. Fix the 400 error, model warning, and context window.
-
The Agentic AI Handbook: Production-Ready Patterns
A comprehensive guide to 113 production-informed patterns for building reliable AI agents.
-
Agent Labs Are Eating the Software World
Why product-first AI startups will dominate the next decade while model labs build the infrastructure they run on
Small tools. Specific problems.
Wire ↗Turn a browser task into a recorded run and a reusable script.skillctl ↗Manage skills across coding agents. Install, sync, and inspect what each agent uses.Durable Researcher ↗Research the web, save progress, and resume interrupted work to produce sourced reports.An ongoing practice
Building, testing, and writing in public. My now page keeps track of what has my attention.
See what I’m working on ↗