Patterns
Workbench, tool, capability, operation, vault - all the same holon shape. These pages are the building blocks: how each holon type is defined, what it composes, and the patterns that recur when you build with it. Read Concepts first if you haven't - these assume you know what a holon is.
Capability
The deterministic behavioral holon - typed input, typed output, optional guardrails. Simple transforms, services from IoC, output hooks/handles, composition, and agents inside a Capability.
Operation
The strategy holon - composes Capabilities, Agents, and sub-Operations into a named, runnable action. The three ways to run one.
Workflow
The long-running holon - declarative .Flow() graphs or imperative .Execute(),
approval gates, parallel steps, and durable persistence.
Agent
The AI behavioral holon - a typed input and a message-based handle. Services, spawning external tools, self-governance, composition, install.
Governance
The guardrail holon - trust levels, allowlists, custom logic, and composites. A constraint that's part of an entity, not a gate in front of it.
Composition
How the holons fit together - the hierarchy, vertical vs horizontal composition, the governance cascade, and a decision tree for picking the right entity.
To build your own from scratch, the Guides walk through creating and testing holons step by step.
On this page
- Patterns
- ╰─▶Capability
- ╰─▶Operation
- ╰─▶Workflow
- ╰─▶Agent
- ╰─▶Governance
- ╰─▶Composition