docs: add 21 concept diagrams across core documentation

Generated with GPT Image 2 (ALBEDO atmosphere, Arakawa geometric precision).
All images in documentation/assets/diagrams/.

ndo_prima_materia.md (5 images):
- NDO three-layer model (pay-as-you-grow activation)
- Brain/NDO layer analogy (complexity matching)
- Stigmergic attachment surface (capability slots)
- Unyt coordination overhead pyramid
- Flowsta dual-signed identity (IsSamePersonEntry)

governance-operator-architecture.md (3 images):
- Governance-as-operator boundary (data vs logic separation)
- Governance evaluation pipeline (4-stage pure function)
- Full cross-zome request sequence (swimlane)

ndo-v1-architecture-design.md (2 images):
- Dual-DNA architecture (hREA + NDO)
- Zome responsibility matrix

architecture_overview.md (3 images):
- Full system topology (Lobby + 3 zomes + hREA)
- PPR category decision tree
- Four-layer privacy model

resources.md (3 images):
- Goods typology 2x2 matrix (excludability x rivalry)
- Property regime transfer rights matrix
- Resource classification to governance defaults

agent.md (5 images):
- Agent type spectrum (Individual to Bot)
- Individual vs Person two-layer identity
- Collective agent dual-face model
- Affiliation spectrum state machine
- Agent as network bridge node

Author: Soushi888