diff --git a/hive-c0re/src/auto_update.rs b/hive-c0re/src/auto_update.rs
index e970fda..51d6e6a 100644
--- a/hive-c0re/src/auto_update.rs
+++ b/hive-c0re/src/auto_update.rs
@@ -78,8 +78,26 @@ pub async fn rebuild_agent(coord: &Arc<Coordinator>, name: &str, current_rev: &s
 /// the approval queue — manager is required infrastructure. Idempotent.
 pub async fn ensure_manager(coord: &Arc<Coordinator>) -> Result<()> {
     let existing = lifecycle::list().await.unwrap_or_default();
+    let current_rev = current_flake_rev(&coord.hyperhive_flake);
     if existing.iter().any(|c| c == MANAGER_NAME) {
-        tracing::debug!("manager container already present");
+        // Container exists already. If it predates the unified lifecycle
+        // (no applied flake on disk) we must rebuild — otherwise it's
+        // running whatever the host-declarative config was at create
+        // time, with a wrong systemd unit and port.
+        let applied_flake = Coordinator::agent_applied_dir(MANAGER_NAME).join("flake.nix");
+        if !applied_flake.exists()
+            && let Some(rev) = current_rev.as_ref()
+        {
+            tracing::warn!(
+                "manager container exists but no applied flake — forcing rebuild to migrate"
+            );
+            let coord_clone = coord.clone();
+            if let Err(e) = rebuild_agent(&coord_clone, MANAGER_NAME, rev).await {
+                tracing::warn!(error = ?e, "manager migration rebuild failed");
+            }
+        } else {
+            tracing::debug!("manager container already present");
+        }
         return Ok(());
     }
     tracing::info!("manager container missing — spawning");
@@ -96,7 +114,7 @@ pub async fn ensure_manager(coord: &Arc<Coordinator>) -> Result<()> {
         &claude_dir,
     )
     .await?;
-    if let Some(rev) = current_flake_rev(&coord.hyperhive_flake) {
+    if let Some(rev) = current_rev {
         let _ = std::fs::write(rev_marker_path(MANAGER_NAME), rev);
     }
     Ok(())