Defining properties
A proposed set of organizational properties and the evidence needed to recognize them.
#Work and resources
Two further properties concern action: discretionary organization of work and access to resources adequate for that work. The organization can select or coordinate relevant tasks rather than merely repeat a fixed schedule. Resources can include compute, documents, tools, budgets, or access supplied by another party; owning money is not essential.
Actual capabilities depend on runtime permissions. A written instruction cannot grant a filesystem permission or a service account privilege. A useful description identifies both the intended boundary and the enforced boundary, together with what happens when an apparently permitted action is unavailable.
#Continuity and outcome assessment
The final proposed properties are durable obligations and assessment of outcomes. A system needs to preserve enough state to resume, resolve unfinished work, and avoid treating every restart as a new beginning. It also needs a way to determine whether an attempted action achieved the intended result.
These are architectural properties, not guaranteed performance. To recognize them, inspect a completed case and an interrupted case: which obligation persisted, which evidence supported acceptance, and which uncertainty remained unresolved? Independent review, multiple agents, databases, and particular message protocols are possible mechanisms rather than universal requirements.
Sources & context
- Wooldridge, Jennings & Kinny (2000). The Gaia Methodology for Agent-Oriented Analysis and Design
- Hübner, Sichman & Boissier (2002). A Model for the Structural, Functional, and Deontic Specification of Organizations in Multiagent Systems
Reference edition 0.1 · 2026-10-14. Read the editorial method.