fix: 部署脚本 worker health-cmd 中 $pid 未转义导致 staging 部署失败回滚

ci_staging_deploy.sh 开启 set -eu，worker 容器 --health-cmd 双引号内的
$pid 被外层部署脚本提前展开（pid 未定义）→ 'pid: parameter not set' →
worker 容器 docker run 失败 → 判定部署失败并自动回滚。

将 health-cmd 内 $pid 转义为 \$pid，使其透传到容器内由 healthcheck shell
展开（for pid in /proc/[0-9]*/cmdline）。回滚段(195行)与主部署段(504行)
两处一并修复。

🤖 Generated with [Claude Code](https://claude.com/claude-code)

scripts/ci_staging_deploy.sh

@@ -192,7 +192,7 @@ rollback() {
       -e PUBLIC_API_BASE_URL=https://staging-api.xiaoxiajianji.com \
       -v "$GENERATED_DIR:/app/generated" \
       --restart unless-stopped \
-      --health-cmd "sh -c \"for pid in /proc/[0-9]*/cmdline; do if grep -ql celery \"$pid\" 2>/dev/null; then exit 0; fi; done; exit 1\"" \
+      --health-cmd "sh -c \"for pid in /proc/[0-9]*/cmdline; do if grep -ql celery \"\$pid\" 2>/dev/null; then exit 0; fi; done; exit 1\"" \
       --health-interval 30s \
       --health-timeout 10s \
       --health-retries 3 \
@@ -501,7 +501,7 @@ docker run -d \
   -e PUBLIC_API_BASE_URL=https://staging-api.xiaoxiajianji.com \
   -v "$GENERATED_DIR:/app/generated" \
   --restart unless-stopped \
-  --health-cmd "sh -c \"for pid in /proc/[0-9]*/cmdline; do if grep -ql celery \"$pid\" 2>/dev/null; then exit 0; fi; done; exit 1\"" \
+  --health-cmd "sh -c \"for pid in /proc/[0-9]*/cmdline; do if grep -ql celery \"\$pid\" 2>/dev/null; then exit 0; fi; done; exit 1\"" \
   --health-interval 30s \
   --health-timeout 10s \
   --health-retries 3 \