Liner notes
Dispatches on AI agents, deep work, and how time actually gets allocated when the machines stop asking for your permission.
Why calendar grids, scheduling links, and auto-decline rules all miss the point. The case for agent-first calendar management.
A technical guide to request classification, model mapping, and provider abstraction. With real cost and latency numbers.
Proactive block creation, defensive alerts, active rescheduling, and fragmentation scoring. A deep dive into protecting deep work.
Why intersection-based scheduling fails and how a negotiation protocol with proposal rounds handles the real-world complexity.
A four-level trust gradient, migration patterns, per-action overrides, and design principles for the hardest problem in AI agents.