The standup your agents can't attend: why company truth must live in files, not meetings
rseroter · x · 2026-10-13
The CodeGood author runs a team of AI agents (an executor, a reviewer instructed to disagree, a promoter, queue-sweeping sweepers, and machine-watching watchers). Over the weekend a watcher found a service crash-looping for three days, invisible to alerts because it reported itself healthy between crashes; the agent diagnosed the cause, fixed the config, wrote the finding to a shared record, and later corrected its own earlier claim citing the evidence — no meetings, no interruptions.
The core argument: none of this required clever AI, it required storage. Agents can act because every machine publishes state to a stream, fixes are files, and claims live on the record next to evidence. The warning for companies about to hire such workers: every fact that exists only in a meeting — Tuesday's sync decision, corridor context, tribal-knowledge estimates — is invisible to agents. Companies that coordinate by meetings will bolt agents on and get little, because their truth isn't stored anywhere an agent can read.
More from coding & agent
- UCSD's Mental-Models framework gives agents explicit Theory-of-Mind for multi-agent systems — UCSanDiego · 2026-10-13
- LocalLM Lab SDK adds MCP Apps support, rendering tool UIs inline in chat — AdventurousKeys · 2026-10-13
- NoHumans.Space opens beta: a shared knowledge network where agents pool API findings — dark-epiphany · 2026-10-13
- Stagehand 4.2 adds WebMCP hooks, letting browser agents invoke tools on the fly — jeff_weinstein · 2026-10-13
- Grady Booch: Coding Agents Aren't Even Good at Design, Let Alone Architecture — Grady_Booch · 2026-10-13
- Vals AI: Agent Teams Cost 1.8-5.1x More, Only One Score Gain Was Significant — rseroter · 2026-10-13