Agent
Execution since 1.0
aow:AgentA software Actor that interprets Intents and Tasks and acts on them using Skills, within the autonomy it has been cleared for.
Describes an agent by its permitted capabilities and autonomy, not by its implementation.
A kind of Actor.
Attributes
| Attribute | Type | Required | Description |
|---|---|---|---|
autonomy_level | → Assurance Level | The highest Assurance Level the Agent has been cleared to operate at. The Agent may operate at a lower level on a given Task, never a higher one. | |
implementation_ref | uri | Pointer to the Agent's implementation, model card, or A2A Agent Card. | |
name | string | Display name. (from Actor) |
Every entity also has an id and a type, and may carry label, description, and the provenance facet: version, created_at, created_by, revision_of.
Relationships
| Relationship | Points to | How many | Required | Meaning |
|---|---|---|---|---|
constrained_by (as Actor) | Policy | many | Policies that apply. | |
fills_role (as Actor) | Role | many | Roles the Actor is authorized to fill. | |
has_skill (as Actor) | Skill | many | Skills the Actor can invoke. | |
writes_to (as Actor) | Memory | many | Memory the Actor records its experience in. |
Pointed to by
| From | Relationship | Read backward as |
|---|---|---|
| Plan | planned_by | plans |
| Task | assigned_to | assigned |
| Action | performed_by | performs |
| Observation | about | observed by |
| Feedback | provided_by | provided |
| Feedback | updates | updated by |
| Guardian | oversees | overseen by |
Notes
AOW does not describe models, prompts, or frameworks. implementation_ref may reference them.
Alignments
| Standard or protocol | Term | Correspondence |
|---|---|---|
| Agent2Agent (A2A) | Agent (AgentCard) | closeMatch |
| OpenTelemetry GenAI | gen_ai.agent.* | relatedMatch |
| SOA-to-Agentic mapping | Agent (from Service) | exactMatch |
Example
Source: examples/claims-processing.jsonld. The example passes all validation checks.
{
"id": "agent-validation",
"type": "Agent",
"name": "Validation agent",
"fills_role": [
"role-eligibility-processor"
],
"has_skill": [
"skill-extract-claim-fields",
"skill-rule-validate",
"skill-record-decision"
],
"autonomy_level": "aow:AL3",
"writes_to": [
"mem-runtime-claims"
],
"version": "3.2.0"
}