Concepts and terminology

Concepts explain the ideas; terminology gives them names. The product pages explain what each project does. These pages explain the vocabulary you need to reason about the system.

Foundations

Page	What it answers
Agency and control	What is agency, autonomy, control, and meta-control?
Governed autonomy	What makes autonomy governable?
Intent and specification	How do goal, context, constraints, and acceptance differ?

Systems and architecture

Page	What it answers
Contracts, not a kernel	Why independent products compose through contracts
Open surfaces	What “open” means when a runtime has commercial editions
Coordination vs orchestration	MCP vs coordination protocol vs durable workflow

Work and capabilities

Page	What it answers
Capability planning	How goals become plans, capabilities, execution, and verification
Capability, skill, and tool	The difference between a tool, a capability, a skill, and an agent
Agent taxonomy	Application, service, employee, physical twin, human twin

Knowledge and context

Page	What it answers
Memory and context	Context, memory, grounding, and world state
World models	Operational world models, digital twins, and possible worlds

Trust and security

Page	What it answers
Agent identity	Issuance, registry, and authorization
Security	The threat model and the controls
Evidence and auditability	What counts as evidence
Observability	Metrics, logs, traces vs evidence

Operations and maturity

Page	What it answers
Reliability	Failure domains, durability, idempotency, recovery
Governance maturity	A practical ladder from ad hoc to adaptive
Anti-patterns	The recurring mistakes

Reference

• Terminology and glossary — the full vocabulary.
• Relationships — the project edge list.
• Reading list — sources and further reading.

These pages are explanatory. They do not override a product’s own documentation.