Skip to content

Concepts

The argument the rest of the project is downstream of. None of these pages tell you how to run anything.

  • Why a graph: Why capability selection is a relationship problem, and where a graph does not help.
  • The harness: Identity, action boundaries, spend, erasure — and what is not built.
  • Surfaces: One graph, more than one way in. Direction, not documentation.

Read them in that order if you are reading them at all: the first says what the problem is, the second what a system owes you once it can act on your behalf, and the third is the part that does not exist and is honest about it.