fix(self-sovereign-cutover): set HR timeout 15m + lower hook deadlines below it (#127)

Two prior provisions (#12 d22b6d4dada2aef2, #14 12e194090631a885) wedged
identically at phase1-watching: bp-self-sovereign-cutover@0.1.25 post-install
hook timed out at 5m (Helm default), Flux marked release Failed, retried 3x,
gave up. catalyst-api never received kubeconfig PUT-back because the cutover
chain inside the Sovereign couldn't complete.

Root cause: HelmRelease had no explicit install/upgrade timeout → Helm's 5m
default → hit before the auto-trigger Job's activeDeadlineSeconds (600s) and
WAIT_TIMEOUT_SECONDS (300s) could complete cleanly.

Fix:
- HR install/upgrade timeout: 15m (covers cold-start cluster + auto-trigger)
- values.autoWaitForAPISeconds: 300 → 720 (12m wait, exits 0 below 15m HR cap)
- values.autoTimeoutSeconds: 600 → 840 (14m Job deadline, below 15m HR cap)
- Chart bump 0.1.25 → 0.1.26

Per CLAUDE.md principle 16: canonical seam = HR timeout + chart hook deadlines.
Both must align: hook deadlines < HR timeout, otherwise Helm gives up before
hook completes regardless of how the Job exits.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

clusters/_template/bootstrap-kit/06a-bp-self-sovereign-cutover.yaml

@@ -241,10 +241,12 @@ spec:
         namespace: flux-system
   install:
     disableWait: true
+    timeout: 15m
     remediation:
       retries: 3
   upgrade:
     disableWait: true
+    timeout: 15m
     remediation:
       retries: 3
   # Per-Sovereign overrides — the chart's values.yaml carries

platform/self-sovereign-cutover/chart/Chart.yaml

@@ -1,6 +1,6 @@
 apiVersion: v2
 name: bp-self-sovereign-cutover
-version: 0.1.25
+version: 0.1.26
 description: |
   Catalyst Self-Sovereignty Cutover Blueprint. Installs DORMANT — this
   chart ships eight step ConfigMaps (PodSpec ConfigMaps, one per step),

platform/self-sovereign-cutover/chart/values.yaml

@@ -331,14 +331,19 @@ trigger:
   catalystAPIURL: "http://catalyst-api.catalyst-system.svc.cluster.local:8080"
   # How long the auto-trigger Job will wait for catalyst-api to be
   # reachable before giving up (and exiting 0 so the operator can fire
-  # manually). 5 minutes is enough for a Sovereign mid-cold-start.
-  autoWaitForAPISeconds: 300
+  # manually). Must finish below the HelmRelease install/upgrade
+  # timeout (15m for bp-self-sovereign-cutover) AND the activeDeadline
+  # below so the Job exits cleanly even when catalyst-api never comes
+  # up — 12 minutes leaves a healthy 3m buffer below the 15m HR cap.
+  autoWaitForAPISeconds: 720
   # Overall cap on the auto-trigger Job runtime. activeDeadlineSeconds
   # on the Job spec — anything longer means catalyst-api is sick and
   # the operator should investigate. The Job exiting at this deadline
   # is non-fatal for the chart install (the cutover engine already
   # runs detached inside catalyst-api once /start returns 200).
-  autoTimeoutSeconds: 600
+  # Must stay below the HelmRelease install/upgrade timeout (15m =
+  # 900s) so the Job ends and the hook unblocks before Helm gives up.
+  autoTimeoutSeconds: 840
   # TTL on the completed Job — kept for audit so operators can read
   # the trigger Pod logs if something looks wrong.
   autoJobTTLSeconds: 86400