fix(tui): don't clear conversation model override on transient retrieve errors

The sync effect's catch block was calling applyAgentModelLocally() on any
conversations.retrieve failure, which cleared the hasConversationModelOverride
flag and reset the model display to agent defaults — even though the override
was still persisted in the DB. This caused intermittent model snap-back that
didn't require a recent /model command to trigger: any transient network error
during the sync cycle would discard the local override knowledge, and without
another agentState change to re-trigger the effect, the reset persisted for
the remainder of the turn.

The fix simply preserves current local state on error. The next sync cycle
will retry the retrieve and self-correct if needed.

👾 Generated with [Letta Code](https://letta.com)

Co-Authored-By: Letta Code <noreply@letta.com>

Author: jnjpng

src/cli/App.tsx

@@ -3376,7 +3376,13 @@ export default function App({
           "Failed to sync conversation model override: %O",
           error,
         );
-        applyAgentModelLocally();
+        // Preserve current local state on transient errors — the override flag
+        // was set by a successful /model write and should not be cleared by a
+        // failed read. The next sync cycle will retry and self-correct.
+        debugLog(
+          "conversation-model",
+          "Keeping current model state after sync error (override in DB is authoritative)",
+        );
       }
     };