feat(system): process heartbeat + /system/health, engine runs business fields (W9-I)

- process_heartbeat / process_heartbeat_sample (0034): api / worker / worker:agent / all
  upsert every 15s (event loop lag last/max, concurrency, running runs, SSE connections, memory)
- GET /system/health (admin): processes with 45s stale flag, SSE, DB latency/size/connections,
  disks (DATA_DIR / library roots / download dirs, cache thresholds), mount sentinel checks
  with IO timeout, engine pending/oldest/24h failure rate, schedules
- GET /system/health/history?hours=: bucketed lag / running / SSE series
- /engine/runs and /engine/runs/{id}: business {kind, refId, title, triggeredBy} joined from
  run input / task / agent_run / tasks_run_meta / outbox; businessKind / businessRefId filters
- docs/runbook/ops.md: field meanings and alert thresholds

Co-Authored-By: Claude Opus 5.5 <[email protected]>

packages/db/src/types.ts

@@ -194,6 +194,7 @@ export type PolicyRevisionStatus = "draft" | "published" | "superseded" | "rolle
 export type PreferenceTagKind = "soft" | "hard" | "version" | "target";
 export type PreferenceTagOrigin = "manual" | "conversation" | "agent_inferred";
 export type PreferenceTagStatus = "active" | "disabled" | "deleted";
+export type ProcessHeartbeatProcessKind = "api" | "worker" | "worker:agent" | "all";
 export type ReconcileRunKind = "identity" | "metadata" | "asset" | "acquisition" | "server";
 export type ReconcileRunMode = "normal" | "read_only";
 export type ReconcileRunStatus = "queued" | "running" | "succeeded" | "partial" | "failed" | "cancelled";
@@ -1079,6 +1080,32 @@ export interface PreferenceTagTable {
   updated_at: GeneratedTimestamptz;
 }
 
+export interface ProcessHeartbeatTable {
+  instance_id: string;
+  process_kind: ProcessHeartbeatProcessKind;
+  pid: number;
+  host: string;
+  version: string;
+  started_at: Timestamptz;
+  last_beat_at: GeneratedTimestamptz;
+  event_loop_lag_ms: Generated<number>;
+  event_loop_lag_max_ms: Generated<number>;
+  concurrency: Generated<Json>;
+  running_runs: number | null;
+  extra: Generated<Json>;
+}
+
+export interface ProcessHeartbeatSampleTable {
+  id: Generated<number>;
+  instance_id: string;
+  process_kind: string;
+  at: GeneratedTimestamptz;
+  event_loop_lag_ms: Generated<number>;
+  event_loop_lag_max_ms: Generated<number>;
+  running_runs: number | null;
+  sse_connections: number | null;
+}
+
 export interface ReconcileRunTable {
   id: Generated<string>;
   kind: ReconcileRunKind;
@@ -1558,6 +1585,8 @@ export interface DB {
   policy_revision: PolicyRevisionTable;
   preference_suppression: PreferenceSuppressionTable;
   preference_tag: PreferenceTagTable;
+  process_heartbeat: ProcessHeartbeatTable;
+  process_heartbeat_sample: ProcessHeartbeatSampleTable;
   reconcile_run: ReconcileRunTable;
   resolution_attempt: ResolutionAttemptTable;
   rule_atom: RuleAtomTable;
@@ -1723,6 +1752,12 @@ export type PreferenceSuppressionUpdate = Updateable<PreferenceSuppressionTable>
 export type PreferenceTagRow = Selectable<PreferenceTagTable>;
 export type NewPreferenceTag = Insertable<PreferenceTagTable>;
 export type PreferenceTagUpdate = Updateable<PreferenceTagTable>;
+export type ProcessHeartbeatRow = Selectable<ProcessHeartbeatTable>;
+export type NewProcessHeartbeat = Insertable<ProcessHeartbeatTable>;
+export type ProcessHeartbeatUpdate = Updateable<ProcessHeartbeatTable>;
+export type ProcessHeartbeatSampleRow = Selectable<ProcessHeartbeatSampleTable>;
+export type NewProcessHeartbeatSample = Insertable<ProcessHeartbeatSampleTable>;
+export type ProcessHeartbeatSampleUpdate = Updateable<ProcessHeartbeatSampleTable>;
 export type ReconcileRunRow = Selectable<ReconcileRunTable>;
 export type NewReconcileRun = Insertable<ReconcileRunTable>;
 export type ReconcileRunUpdate = Updateable<ReconcileRunTable>;