The rearchitecture that made Managed Agents work: split the 'brain' (model plus agent harnessThe scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.Full definition →) from the 'hands' (sandboxes and tools) and the 'session' (the event log), each an interface that assumes almost nothing about the others.
Planning, not coding, was the biggest phase of a six-month AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition →-infra project, and the artifact was still a Google Doc PRD — because its job is aligning the sandboxAn isolated environment where AI-generated code or agent actions run without being able to touch anything real.Full definition → team, cloud partners, and product, which AI does not speed up.
They de-risked the product by building it for an internal customer first: the backend for Claude Code on the web was the same shape of problem, so it served as the spike before generalizing it for outside customers.
Negotiating interfaces between teams got cheaper. Instead of handing over a fully spec'd interface, they stood up a stub service that shadowed traffic and settled the contract while both sides kept building.
Anthropic's API layer is migrating from Python to Rust. Python was picked for researcher familiarity and fast iteration, but it is single-threaded and does not hold up on the token hot path under load.
Terms in this piece · Glossary
agent harness — The scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
sandbox — An isolated environment where AI-generated code or agent actions run without being able to touch anything real.
Why it matters
It gives engineers a rare, specific look at how AI-native teams actually sequence work — where planning artifacts still pay off, how far parallel agent fan-out can compress a year-long rewrite, and why verification is now the expensive step. Useful as a benchmarkA standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.Full definition → for restructuring your own agent workflow rather than guessing from vendor marketing.
Key quotes
“Historically, rewrites are a terrible idea.”
Jarred Sumner
“We see that few tokens are spent on actual implementation. Most are spent on discovery of unknowns, prototyping, mocking, and then in verification and testing.”
