Concepts

A small set of signature concepts. Each one names something the agent-engineering conversation kept referring to without a shared word. Each has a permanent URL, so an essay or talk can point at the concept rather than re-define it.

• Four Layers: What Each Part of This Work Is Actually For

  Cases are reference material. Concepts are the cognitive foundation. Patterns are the common language. The dual-axis matrix is the container. The meta-essay; read this one first.

• Model Spends, Harness Budgets

  Agent architecture is the design of bounded resource allocation under uncertainty. The interesting work is in the harness, not the prompt.

• Cognitive Function × Execution Topology

  Two orthogonal axes. What the agent is doing, and what shape the work takes. Crossed, they form the matrix from which 28 named patterns are read.

• Pattern Coordinates

  A pattern name without coordinates is a tag, not a design decision. The coordinate is what makes a choice testable.

• Cost as Design Constraint

  When the model is fixed and cost is the binding constraint, pattern selection IS the design. This is where engineering judgment lives.

• Choreography Is Not a Topology

  Choreography is a coordination mode that operates across topologies, not one column among six. Closer to the Teams pattern than to Orchestrate's opposite.

New concepts are added when the same idea has surfaced in three or more pieces of writing without a name. The bar is high on purpose.