Agent workflows shifted in 6 months: from babysitting 6 windows to one 18-hour autonomous session
gregmushen · x · 2026-10-13
Developer Greg Mushen documents a six-month shift in AI coding agent workflows: in June, short runs requiring heavy human intervention were the norm; now long-running autonomous sessions are standard.
He used to monitor six agent windows in real time — now he casually checks a single window while doing other things. The attached screenshot shows a session running 18hr 41m on Opus 5.5, 91% context used, $755.84 in cost, with multiple background sub-agents handling DB tests, bug tracing, and script writing in 19–52 minute stretches.
More from coding & agent
- Swapping only the harness moves a model 35 points: 8090 Labs tests 7 models across 5 coding agents — pstAsiatech · 2026-10-13
- Meta launches Muse Wonder frontier model family, installable with one curl command — jiayi_pirate · 2026-10-13
- Harness code becomes a research battleground: CMU rewrites it with RL, Apple says it doesn't matter — burkov · 2026-10-13
- FrontierCode 1.1 leaderboard: Claude Opus 5.5 tops at 52% as leading models fail 40%+ of tasks — ArtificialAnlys · 2026-10-13
- Anaconda adds Kilo Agent Swarms and Enkrypt AI autonomous red-team agents — anacondainc · 2026-10-13
- Open-source agent workspace Open Dots adds Meta's new Muse Wonder models — matchaman11 · 2026-10-13