Entries tagged with: ARCHITECTURE

What we learned from making a research agent durable, routed, visible, and eval-driven.

Claude Code's deep-research workflow, pulled from its binary and dissected. Wide search, no second hop.

How extensibility shifts from packaged plugins to agent-readable recipes, and why the connector is the new extension point.

Why the gap between AI demos and shipping AI is a reliability gap, not a capability gap.

Files ending in .md trigger automatic processing that breaks agent instruction files. Use dotfiles instead.

How I built a dual-format delivery system serving identical content to humans and AI agents with no hidden restrictions.