The causal backbone of StoneOS
How Vectry Analytics, Veripass and Sommatic compose: identity grounds the events, workflows emit them, and the causal graph explains the whole operation.
Vectry Analytics wasn’t built in a vacuum. It’s the causal backbone of StoneOS — the Blackwood Stone Holdings stack where identity, cognitive operations, commerce, fintech and edge hardware compose into complete systems. That context shaped the product in ways worth explaining.
Identity grounds the graph
Every Vectry event names its actor. In the StoneOS stack, that actor isn’t a free-form string — it’s an identity governed by Veripass, the federation and contextual-access layer. When an event says actor_id: 'usr-99992', there’s a verified identity behind it, with roles, scope and risk posture.
This is why the context field exists on every primitive: platform integrations attach identity context (like the acting username) so the causal graph can answer not just what caused this but who was allowed to.
Workflows emit into the same graph
Sommatic, the cognitive operations platform, runs workflows through 130+ node operators. Five of those operators are published by Vectry Analytics as a verified publisher on the StoneOS Marketplace:
vectry.log— emits a structured causal event from any workflow steptrace.start/trace.end— opens and closes execution tracesmetrics.emit— operational metrics with causal contextaudit.append— append-only audit records
There’s also an event trigger: workflows can subscribe to the Vectry causal event bus and fire when a matching event arrives — filtered by domain, entity, operation or tags. The loop closes: operations emit events, events trigger operations.
Why this matters even outside StoneOS
The integration posture is deliberately non-invasive: systems of record stay authoritative, Vectry observes and threads. That same posture applies to your stack. The HTTP ingest API that Sommatic’s workers use is the same one your services use — one wire contract, whatever emits into it.
But the composition shows what the causal layer is for: when an automated workflow takes an action on a real operation, the thread that explains it — who triggered what, through which steps, ending in which outcome — already exists. Governance isn’t a report you assemble afterwards; it’s a property of the data layer.
That’s the StoneOS thesis applied to data: systems compose when the story composes.