ci(deploy): retry compose up once to survive container-recreate races

A raced or interrupted previous deploy (e.g. a manual compose up overlapping the automated one) could leave a half-removed container behind, so the next recreate failed with 'removal of container ... is already in progress' (seen on the b25eb44 run, idx15). Prune stopped containers and retry once.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

scripts/deploy.sh

@@ -15,7 +15,20 @@ git fetch --all --prune
 git reset --hard origin/main
 
 echo "==> Building + starting the stack…"
+deploy_up() {
   docker compose -f "$COMPOSE_FILE" --env-file "$ENV_FILE" up -d --build --remove-orphans
+}
+
+# A raced or interrupted previous deploy (e.g. a manual `compose up` overlapping
+# the automated one) can leave a half-removed / renamed container behind, making
+# the next recreate fail with "removal of container ... is already in progress".
+# If the first attempt fails, prune stopped containers and retry once.
+if ! deploy_up; then
+  echo "!! 'compose up' failed — pruning stopped containers, retrying once in 8s…"
+  docker container prune -f >/dev/null 2>&1 || true
+  sleep 8
+  deploy_up
+fi
 
 echo "==> Pruning dangling images…"
 docker image prune -f >/dev/null 2>&1 || true