kami/orchestra
1
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases Wiki Activity

Record what an operator repaired, and let debt count incidents

Browse Source 
Slice B, second half. OperatorInterventionRecorded is the one command for
saying "I fixed this by hand": a manual repair happens outside Orchestra by
definition, so the only honest way to have the evidence is for the person
who made it to state it. Inferring "an operator probably intervened" from a
gap would put guesses into the record the ledger is built from.

The debt projection now consumes both new kinds. A closed incident is one
observation carrying its repeat count as intensity, so recurrence stays a
count of independent incidents: 301 repeats on one lease and 2 on another
is a recurrence of two with an intensity of 303, not a recurrence of 303.

Both kinds were previously reported as holes in the system. They are
ordinary evidence now, so their absence from a history is a fact about that
history, and the gap list says so.

The worker also stamps a per-process incarnation on registration and every
heartbeat. Nothing else on the wire distinguishes a restarted worker from a
running one, and an incident cannot outlive the process that reported it.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CVbaKucEYBjMqVeUgJUsc1
This commit is contained in:
kami
2026-08-30 06:53:26 +04:00
parent 438c1d6df3
commit e0601296e0
9 changed files with 337 additions and 29 deletions
Download Patch File Download Diff File Expand all files Collapse all files
internal/domain/domain.go
+5 -1
View File
@@ -305,6 +305,8 @@ func EventWithoutTask(typ string) bool {
switch typ {
case "QuotaReported", "StandupAdvisory", "ApprovalGranted", "ApprovalDenied",
EventObservationIncidentOpened, EventObservationIncidentClosed:
// An intervention is deliberately absent: it names a task when it
// repaired one, and that task must exist.
return true
}
return false
@@ -317,7 +319,7 @@ func ValidateEvent(e Event) error {
if e.SchemaVersion >= 2 && strings.TrimSpace(e.Surface) == "" {
return fmt.Errorf("%w: surface required", ErrInvalid)
}
allowed := map[string]bool{"TaskCreated": true, "TaskLeased": true, "TaskLeaseRenewed": true, "TaskReleased": true, "TaskLaunchAcknowledged": true, "TaskPickupValidated": true, "TaskCompleted": true, "TaskFailed": true, "TaskBlocked": true, "TaskNeedsAttention": true, "ApprovalRequested": true, "ApprovalGranted": true, "ApprovalDenied": true, "TaskAmended": true, "TaskCorrected": true, "QuotaReported": true, "StandupAdvisory": true, EventHumanDecisionRecorded: true, EventHumanDecisionSuperseded: true, EventWorkPhaseChanged: true, EventDeferredFindingRecorded: true, EventReviewRecorded: true, EventTaskSubmitted: true, EventTaskChangesRequested: true, EventPlanPhaseVerified: true, EventPlanMismatchRecorded: true, EventObservationIncidentOpened: true, EventObservationIncidentClosed: true}
allowed := map[string]bool{"TaskCreated": true, "TaskLeased": true, "TaskLeaseRenewed": true, "TaskReleased": true, "TaskLaunchAcknowledged": true, "TaskPickupValidated": true, "TaskCompleted": true, "TaskFailed": true, "TaskBlocked": true, "TaskNeedsAttention": true, "ApprovalRequested": true, "ApprovalGranted": true, "ApprovalDenied": true, "TaskAmended": true, "TaskCorrected": true, "QuotaReported": true, "StandupAdvisory": true, EventHumanDecisionRecorded: true, EventHumanDecisionSuperseded: true, EventWorkPhaseChanged: true, EventDeferredFindingRecorded: true, EventReviewRecorded: true, EventTaskSubmitted: true, EventTaskChangesRequested: true, EventPlanPhaseVerified: true, EventPlanMismatchRecorded: true, EventObservationIncidentOpened: true, EventObservationIncidentClosed: true, EventOperatorInterventionRecorded: true}
if !allowed[e.Type] {
return fmt.Errorf("%w: unknown type %q", ErrInvalid, e.Type)
}
@@ -603,6 +605,8 @@ func ValidatePayload(typ string, p map[string]any) error {
return ValidateObservationIncidentOpened(p)
case EventObservationIncidentClosed:
return ValidateObservationIncidentClosed(p)
case EventOperatorInterventionRecorded:
return ValidateOperatorInterventionRecorded(p)
case EventReviewRecorded:
if err := requiredHash(p, "artifact_ref"); err != nil {
return err