Alert Is Not Decision

An alert — a threshold breach, an anomaly flag, a red tile, a pager page, or a silenced notification — can surface that something crossed a rule. Seeing, acknowledging, or silencing an alert is not a Decision Case and not a decision. Sync may emit alerts, coverage gaps, and recommendations. A named human still opens a Decision Case on a question, cites approved evidence, and accepts, rejects, escalates, or returns. Clearing an alert queue, muting a channel, or auto-closing a ticket must not be read as authorization, verification of outcome, work-order execution, or plant control.

Reliability work collapses the signal because a quiet queue looks like a closed case. Someone acknowledges the page. Someone mutes the channel. A ticket closes itself when the tile returns inside the rule. The meeting treats those acts as if a question had been decided, work had been authorized, an outcome had been checked, or the plant were under direction. They were not. The alert showed that something crossed a rule. No named person had opened a Decision Case on a question, cited approved evidence, and accepted, rejected, escalated, or returned a recommendation.

Sync keeps that split on the signed-in Decision Case. A signed-in user completes the case in a fixed order: Question, Evidence, Recommendation, Human decision, Action, Verification, and Learning. Orville Davis states that order in Field Manual v0. The manuals index lives at /manuals. This essay is why an alert cannot be read as the human decision or as verification. The manual is the public contents. The Human decision chapter and the Verification chapter are the two steps this essay keeps apart from the signal.

An alert surfaces a rule crossing

An alert is a signal that something crossed a rule. A threshold breach says a reading passed a limit someone already set. An anomaly flag says a calculation marked a point as outside the pattern it was written to watch. A red tile paints that condition on a display. A pager page carries the same condition to a person who is not looking at the screen. A silenced notification is still that signal: the rule fired, and the notice was suppressed. Each form can make a rule crossing visible. None of them opens a case, and none of them names who may decide it.

Dashboard Is Not Decision already places the display on the other side of the decision. KPI tiles, trend charts, health scores, and red, amber, and green status can surface plant or maintenance state. Seeing that display is not a Decision Case. This essay is the next boundary. The alert is the signal that a rule fired, including when the signal is a red tile on that dashboard, a page sent from it, or a notification someone has already silenced. The signal is not the case that would decide what to do about it.

A rule crossing is not a question on a case. Acknowledging the signal is not a named decision.

The rule can be current. The flag can match the calculation. The page can reach the person on shift. None of that records a reliability or maintenance question, the evidence cited for it, or a human act on a recommendation.

Seeing, acknowledging, or silencing is not a Decision Case

A Decision Case opens on a reliability or maintenance question a named person can decide. The question is recorded before evidence or recommendation. Question Is Not Decision states that opening. An alert does not write the question. A person does, on the signed-in case.

Seeing the alert is not that opening. Acknowledging it is not that opening. Silencing the notification is not that opening. The signal can fire while no case exists. The queue can clear while an existing case stays where it was. The alert and the case are different records. One shows that a rule was crossed. The other holds a question a named person can decide.

Seeing, acknowledging, or silencing an alert is not a Decision Case and not a decision.

Sync may emit alerts, coverage gaps, and recommendations. Emitting them puts a signal, a gap in what is watched, and a proposed next action where a person can read them. The three can sit on the same screen. They are still not the case. A named human still opens a Decision Case on a question. Until that opening is recorded, Sync has emitted a signal, a gap, or a proposal. It has not started the loop.

The decision is a named human act

The human decision comes after a recommendation exists to decide. A recommendation states a proposed next action, the evidence it uses, and the uncertainty it leaves in view. Drafting that proposal leaves authority where it already sits. Recommend Is Not Authorize states that split. An alert shown next to the draft does not close it. A coverage gap shown next to the draft does not close it either.

A named human still opens the case on a question, cites approved evidence, and accepts, rejects, escalates, or returns. Citing approved evidence means the decision points at the record already on the case: observed fact, assumption, and what is still missing. A threshold breach is not that citation. An anomaly flag is not that citation. The Human decision chapter publishes the act. The case keeps the named person beside the proposal. Until the act is on the case, what remains is a proposal. The person who can accept the operational consequence is still the one who decides.

Until that act is on the case, the queue can change and the decision is still absent. Acknowledging an alert does not accept a recommendation. Silencing a notification does not reject one. Muting a channel does not escalate. Auto-closing a ticket does not return the proposal to the person who drafted it. Those are signal-handling acts. The four decisions are a named person on the case.

A cleared queue is not verification

Verification is the check after a named human has authorized an action. It asks whether that action produced the result the decision named, and it names the evidence on the case that would show it. The signed-in case holds that check next to the question, the proposal, and the person who decided. The Verification chapter publishes that step. Verification Is Not Optional states the gate: the case is not closed until the check is recorded.

Clearing an alert queue, muting a channel, or auto-closing a ticket must not be read as verification of outcome.

A queue can clear because a reading fell back inside the rule, because someone edited the rule, because the channel was muted, or because a ticket timer closed the item. A silenced notification can stay silent for the same kinds of reasons. None of those events records who authorized an action, or whether the authorized action did what it was supposed to do. An alert that was never tied to an authorized action has nothing authorized to check. Citing a cleared queue as if the outcome were known treats a signal state as a closed case.

Stage-1 verification is the check written on the case. This edition does not treat a live pull of historian or control-system tags as that check. An alert fed by those tags, or by simulated or seeded practice records, is still a signal. It is not the verification step. Learning that follows verification inherits the closed case. It does not inherit a queue that went quiet.

Clearing, muting, or auto-closing is not authorization, execution, or control

Clearing an alert queue, muting a channel, or auto-closing a ticket must not be read as authorization, work-order execution, or plant control.

Authorization is the named human decision on a recommendation. The alert sits outside that act. Acknowledging a page does not grant permission to do the work. Muting the channel does not either. The recommendation can still be rejected, escalated, or returned after the queue looks empty.

The same signal does not execute the work. Action Is Not Execution states the later boundary. An ACTION disposition records work intent after a human decision. It does not write a work order, isolate equipment, or change the plant. ACTION stays locked until authorized execution systems write the work order or isolate the equipment. A cleared alert queue does not unlock it. An auto-closed ticket is not a work order.

It also does not control the plant. Sync may emit alerts, coverage gaps, and recommendations. A named human decides. Authorized execution systems write the work order or the isolation. None of that follows from a rule crossing, from silencing the notice, or from a ticket that closed itself. A Decision Case that stores an alert stores a record of a signal. That storage is not permission for Sync to order work, isolate equipment, or change the plant.

Where the public statement lives

Field Manual v0 is the public contents of this loop. Start at the manuals index or open Sync Field Manual directly. Human decision is step 04. Verification is step 06. An alert is neither step. The signal can sit beside the case. It does not replace the named act, and it does not replace the check. Later editions can deepen a chapter. The spine stays in this order.

What this article is not claiming

This is an essay about the Decision Case order, not a customer case study. It names no plant, states no savings figure, and claims no prevented failure. It states no OEM limit and no operating threshold. It does not claim that an alert authorizes work, verifies an outcome, executes a work order, or controls a plant.

Stage-1 readiness means a signed-in user can complete the Decision Case — question, evidence, recommendation, human decision, action, verification, and learning — and Field Manual v0 describes that journey. This edition does not describe plant execute, a live connector tag pull, SMTP invite delivery, or automatic revocation of access on expiry as live. It does not describe Sync writing work orders, isolating equipment, or controlling the plant. Simulated or seeded telemetry and assets are practice records. They are not live plant results. Self-guided onboarding is not claimed as a live product path.

The series continues with Honesty Boundary Is Not Optional, on why Sync states what is simulated, seeded, or out of scope, and why a Decision Case that pretends plant authority Sync does not hold is not a Decision Case. Companion reading: Dashboard Is Not Decision on why a dashboard of plant or maintenance state is not a Decision Case and not a decision, Question Is Not Decision on why recording a reliability or maintenance question is not the human decision, Recommend Is Not Authorize on why a drafted proposal is not permission, Verification Is Not Optional on why the case stays open until the check is recorded, and Action Is Not Execution on why a recorded ACTION disposition is not plant execution. A Reliability Assessment asks whether the records can support a conclusion. A Strategic Pilot is a governed proof around one operating decision. The Human decision chapter is the named act. The Verification chapter is the check. An alert is neither.