PUBLIC ACCESS // THE ARCHIVETHE ARCHIVE
Writing on technology, building, comedy, and the absurdity of everything in between.

What is the best AI classifier? I made them fight to the death to find out.
I'd been looking into which AI classifiers were better for different jobs and writing benchmarks to compare them. Useful work, but I also wanted a way to show people what these models could do....
READ ARTICLE →
One small change will save my AI coding agents hours of waiting each day
I spend quite a lot of time getting AI to write code. It turns out the agents were also spending quite a lot of time waiting for TypeScript. I've been so absorbed in AI that I've neglected some of...
READ ARTICLE →
Claude Code finally supports AGENTS.md, and quietly fixes a much more expensive problem
I shared the quick version of this in today’s video. This is the bit that deserves more than sixty seconds. Anthropic has finally taught Claude Code a filename the rest of the coding-agent world...
READ ARTICLE →
Fast, cheap and confidently wrong: testing Jev on coding-agent safety
Jev had started appearing everywhere in my feeds, usually next to numbers with a lot of xs in them. Fast. Cheap. Built for decisions rather than chat. I was curious. Io was available. This is how...
READ ARTICLE →
Running ten agents broke my brain
A real story about scaling up parallel coding agents, crashing hard, and the tool I built after working out why. !The same engineer later, calm, directing two robot drones at eye level while a...
READ ARTICLE →
A live timestamp in a cached prompt will quietly wreck your bill
Prompt caching only works if the cached part of your prompt never changes. Put anything that changes on every call, even something as innocent-looking as the current time, inside that cached section,...
READ ARTICLE →
io-auto-mode: the missing security layer for autonomous agents
!A gatekeeper figure waves most incoming agents straight through a gate while a second reviewer pulls a few aside for closer inspection Useful and dangerous are the same setting An agent running...
READ ARTICLE →
MCP vs A2A: which one do you actually need?
Both let agentic systems reach further than themselves. Only one of them is worth your attention today. !Illustrated propaganda-poster artwork: one robot figure reaching down toward a service...
READ ARTICLE →
A second opinion only works if it's actually a second opinion
The same read twice If you want a genuine second opinion on your agent's output, it needs to come from a different model, on a different codebase, with different blind spots. That's the whole case...
READ ARTICLE →
We've open sourced the UAIL (Universal AI Level)
Sophie published the Universal AI Level today: a 0 to 10 scale for what someone can actually do with AI, not which tools they have opened. Eleven levels: 0 for never touched it, 10 for building the...
READ ARTICLE →
The Brain: a memory system for my AI agents, and why the "why" is the bit that actually matters
I've built a memory system for my AI agents. I call it the Brain. Here's how it works and why I built it. !Illustration of a lone keeper tending a glowing memory vault, threads of light connecting...
READ ARTICLE →
Spend your expensive model on judgement, not on typing
Fable came back for a short, limited window this month. I burned through 70% of a week's allowance in a single day, because I let it do everything: planning, implementation, verification, testing,...
READ ARTICLE →
MCP or CLI: the right answer depends on where your agent runs
Anthropic invented the Model Context Protocol in late 2024, and it has since become the standard way to let AI agents call external systems. My agents use it every day. They also spend a good part of...
READ ARTICLE →
MCP vs raw API: one interface you learn once, or one you rebuild every time
MCP and a raw API both let an agent reach an external service. Only one of them scales past the second service. I get asked which to use often enough that it is worth writing the answer down once,...
READ ARTICLE →
The Digital Loom: How AI is Democratising Custom Software
When we look back at the Industrial Revolution, we often think of massive, thunderous textile mills. We picture the Luddites protesting against the automation that they feared would replace their...
READ ARTICLE →
The Maker's Mindset: Why Small Victories Lead to Big Outcomes
When you're building a business, it's easy to get caught up in the desire for the finished product. You want the perfectly branded social feed, the comprehensive website, and the marketing strategy...
READ ARTICLE →
Built to Work How You Work: Designing Tech Around Your Life
Have you ever felt like you are working for your tools, rather than the other way around? It is a common trap for small business owners. We find a shiny new piece of software, spend hours trying to...
READ ARTICLE →
How Claude Code injects context: hooks, native memory, and the prompt cache
My agents have long-term memory, and at some point that memory has to actually reach the model. The lazy answer is to shove everything into every prompt, every turn, and let the LLM sort it out. That...
READ ARTICLE →
AI agents have a goldfish memory. Here's how to fix that.
AI agents have a terrible memory. As bad as a goldfish. Past a certain point, they just forget. That's not a criticism, it's a fact you have to design around if you're building an agentic system...
READ ARTICLE →
Why We're Live-Demoting Marketing AI Agents at EVOLVE Brighton Today
If you walk through the doors of the Brighton Centre today, you will likely hear a lot of big, shiny words. Artificial intelligence, paradigm shifts, cognitive architectures, and the inevitable...
READ ARTICLE →
Building an OAuth-ready MCP Server: A Meta Journey
Building an OAuth-ready MCP Server: A Meta Journey There's something wonderfully recursive about writing a blog post using the very tools you're writing about. That's exactly what's happening here:...
READ ARTICLE →