https://thrulinehq.com/blog Research and field notes on AI coding agents and what their “done” rests on. en https://thrulinehq.com/research/you-are-the-latency https://thrulinehq.com/research/you-are-the-latency Thu, 15 Oct 2026 00:00:00 +0000 An essay on meat proxies: what becomes of a person whose job is carrying text from one AI model to another, and why the one job left in the loop is deciding. https://thrulinehq.com/research/we-ran-thruline-on-itself https://thrulinehq.com/research/we-ran-thruline-on-itself Wed, 14 Oct 2026 00:00:00 +0000 AI agents wrote most of ThruLine’s code, and ThruLine recorded them. What its own record caught, what it got wrong about itself, and one fix, end to end. https://thrulinehq.com/research/green-is-not-proof https://thrulinehq.com/research/green-is-not-proof Fri, 09 Oct 2026 00:00:00 +0000 CI checks the code an AI agent leaves behind, not what it did or whether its “done” is true. The evidence from 2025–2026, and what a record changes.