Three-Layer Architecture
The Foundation, Skills, Agents model that structures Knowledge OS into context (Layer 1), repeatable processes (Layer 2), and autonomous workflows (Layer 3).
Three-layer architecture is the organizing principle of Knowledge OS. Layer 1 (Foundation) includes CLAUDE.md, rule files, context files, and the knowledge graph, the persistent memory that makes every session context-aware. Layer 2 (Skills) encodes repeatable processes as SKILL.md files with methodology, references, and quality gates. Layer 3 (Agents) adds autonomy: scheduled execution, multi-agent orchestration, and event-driven triggers. Each layer depends on the one below it. Agents invoke skills, and skills load foundation context. You can run a productive system with just Layer 1, add Layer 2 when you have repeatable workflows, and add Layer 3 when you want things to happen without you.
Where it shows up:
Related Terms
Foundation
Knowledge OS
A persistent, file-based operating system for AI that turns Claude Code into a context-aware busines...
Foundation
CLAUDE.md
The root instruction file that Claude Code reads at session start, containing your business context,...
Skills Layer
Skill
A reusable instruction set that encodes a repeatable process. Skills load context, apply methodology...
Agent Layer
Agent
A skill with autonomy. Agents run on schedules or triggers, execute multi-step workflows, and produc...