Activation

Activation is the bridge between an inbound Signal and durable agent work. It covers the Interpret and Resolve steps of Phase 1 — classifying the Signal, applying Policy, deciding among the five Resolve outcomes, and, when work is warranted, producing an Activation record that names the responsible actor and pins the context the Plan step will consume. Everything downstream in matic — Tasks, Work Plans, Runs, Deliveries — is traceable back to an Activation, so the records produced here are the system's authoritative answer to "why is this agent doing this work?"

Interpret and Triage

Step 2 of the lifecycle: Signal type classification, Policy evaluation, Memory retrieval, priority triage, and routing target identification. Learn more

Resolve Outcomes

The five terminal outcomes of Resolve — Ignore, Reply, Clarify, Activate, Notify — and the rules that select between them. Learn more

Clarify Loop

How matic holds a Signal in conversational Clarify state when activation blockers exist (missing context, ambiguous scope, unmet HITL gate) and how the loop exits into Activate. Learn more

Decision on Activate

The Decision record that is mandatorily emitted for every Activate outcome — what it captures, where it is stored, and how it links Signal, Activation, and assigned actor. Learn more

Group Resolution

How a Signal addressed to a Group is reduced to a single assigned actor, and how that selection is recorded on the Activation alongside the originating Group reference. Learn more

Activation Record

The on-disk shape of an Activation: interpreted Signal reference, triage results, identified actor, gathered context, and the handoff payload consumed by the Plan step. Learn more