Wez Pyke

i have like 4 agent PRs open that i keep not reviewing. turns out thats everyone harvard looked at 718 companies using coding agents. 30% more code, 23% more PRs, finished work basically flat. review time up 49%, almost 2x the PRs sent back who on your team is eating all this

Anthropic is now recommending that companies REMOVE unnecessary subagents from their AI systems. This is a pretty significant shift in agent architecture, with their team finding that a single agent equipped with skills can frequently outperform complex multi-agent systems, and one customer seeing roughly 2x faster execution. The emerging rules of thumb: → Skills: Teach the agent HOW to do something, including procedures, expertise, and instructions. → Tools / MCP: Give the agent the ability to DO things through APIs, actions, and live data. → Subagents: Use for parallel work, isolated context, or tasks a single agent struggles with. The interesting part is progressive disclosure. Instead of spawning specialized agents or stuffing everything into a giant prompt, one agent discovers and loads skills and tools as needed. Their advice is to fold existing subagents into the main loop, turn their instructions into skills, expose their actions as tools, and run evals to see if performance holds up. As models have gotten smarter, improving agent architecture is becoming less about what you can add and more about what you can delete. claude.com/resources/article…

Back to @firefox for me, this is insane

still no mention of the vulnerability on @arcinternet or @browsercompany or anywhere that isn't release notes (which were updated retrospectively)

found out today that credit card numbers are validated by an algo called "Luhn's Algorithm"

lord give me the strength to not overengineer this app for 0 users

Here’s the thing… 🤐 Most documentation tools suck. They’re like one-size-fits-all t-shirts when your brand needs something custom-made. Why is Stripe-level documentation so rare? Every scaling company ends up reinventing the wheel after a few years with cookie-cutter hosted docs tools. Building an API reference with OpenAPI specs? It’s a pain. Tools like Swagger UI are outdated and clunky. It doesn’t have to be this way. Docs infrastructure should be decoupled from the frontend—and offered as SaaS.

Here's the core idea: 1. You get any success (example: first 100 users) 2. You tell people how you got this success (this drives traffic) 3. You tell people how you telling people about your success drove even more traffic (which drives more traffic)