diff --git a/hive-c0re/src/agent_server.rs b/hive-c0re/src/agent_server.rs
index 182e3413..9dcf468e 100644
--- a/hive-c0re/src/agent_server.rs
+++ b/hive-c0re/src/agent_server.rs
@@ -563,6 +563,31 @@ pub(crate) fn handle_send(
     // learning the label — runtime reparenting propagates for free. See
     // `docs/conventions.md::Recipient sentinels`.
     let resolved = crate::topology::resolve_recipient(agent, to);
+    // Validate that the resolved recipient is a known local agent or the
+    // special "operator" recipient. Without this check a typo in `to`
+    // silently queues a message nobody will ever read (issue #1165).
+    //
+    // Cross-hive messaging (`name@hive` qualified names) is not routed
+    // through the broker — use the Matrix MCP tools for that instead.
+    if resolved.contains('@') {
+        return AgentResponse::Err {
+            message: format!(
+                "send failed: cross-hive recipient `{resolved}` is not supported \
+                 via the broker — use Matrix MCP tools for cross-hive messaging"
+            ),
+        };
+    }
+    if resolved != hive_sh4re::OPERATOR_RECIPIENT {
+        let state_root = crate::coordinator::Coordinator::agent_state_root(&resolved);
+        if !state_root.exists() {
+            return AgentResponse::Err {
+                message: format!(
+                    "send failed: unknown recipient `{resolved}` \
+                     (no agent with that name exists on this hive)"
+                ),
+            };
+        }
+    }
     match coord.broker.send(&Message {
         from: agent.to_owned(),
         to: resolved,
diff --git a/hive-c0re/src/lifecycle.rs b/hive-c0re/src/lifecycle.rs
index 4441d0f4..cd4c0555 100644
--- a/hive-c0re/src/lifecycle.rs
+++ b/hive-c0re/src/lifecycle.rs
@@ -467,7 +467,26 @@ pub async fn rebuild_no_meta(
             priv_run("stop", name).await?;
         }
         on_step("nixos-container update");
-        priv_run("update", name).await?;
+        let update_result = priv_run("update", name).await;
+        if let Err(ref update_err) = update_result {
+            // The update failed (e.g. nix build error). If the agent was
+            // running before we stopped it, try to bring it back up on the
+            // previous successful configuration so it doesn't stay dead.
+            // The start failure is logged but not promoted to an error —
+            // we always propagate the original update error (below).
+            if was_running {
+                tracing::warn!(
+                    %name,
+                    error = %update_err,
+                    "nixos-container update failed; attempting restart on old config"
+                );
+                on_step("nixos-container start (recovery)");
+                if let Err(e) = priv_run("start", name).await {
+                    tracing::warn!(%name, error = %e, "recovery start after failed update also failed");
+                }
+            }
+        }
+        update_result?;
         if was_running {
             // Cold-start fallback on activation errors.
             // See `docs/coordinator.md::Cold-start fallback`.