perf(tui): render streaming token secara inkremental + kurangi redraw sia-sia

- fix(view): token streaming kini benar-benar tampil — sebelumnya cache
  display_lines tidak pernah di-rebuild saat pesan terakhir berubah
  (msg_count == cached_count), jadi teks AI streaming tidak pernah muncul
  sampai pesan baru/resize
- perf(view): streaming kini hanya re-render pesan TERAKHIR (splice di
  batas cached_last_start) → O(konten baru) per token, bukan O(seluruh
  history); guard cached_last_len mencegah re-render pada frame spinner
  tanpa token baru
- perf(run): skip chrono::Utc::now() + drain toasts saat tidak ada toast
- perf(misc): drain_expired_toasts tidak lagi clone seluruh daftar toast
- perf(view): render_toasts fast-path saat toasts kosong

apps/interfaces/tui/src/view/chat.rs

@@ -190,6 +190,8 @@ impl Component for ChatComponent {
             }
             state.display_lines_cache = all_lines;
             state.cached_msg_count = msg_count;
+            state.cached_last_start = state.display_lines_cache.len();
+            state.cached_last_len = last_msg_len(state);
             state.render_width_at_cache = state.last_render_width;
             state.transcript_cache.dirty = false;
         } else if state.transcript_cache.dirty && msg_count > cached_count {
@@ -205,6 +207,27 @@ impl Component for ChatComponent {
                 state.display_lines_cache.extend(msg_lines);
             }
             state.cached_msg_count = msg_count;
+            state.cached_last_start = state.display_lines_cache.len();
+            state.cached_last_len = last_msg_len(state);
+            state.transcript_cache.dirty = false;
+        } else if state.transcript_cache.dirty {
+            // Streaming update: message count unchanged, but the last
+            // assistant message's content grew (StreamToken). Re-render only
+            // that message and splice its lines back into the cache — this
+            // keeps updates O(new content), not O(whole history) per token.
+            // Without this branch, streamed tokens would never appear (the
+            // cache is only rebuilt on new-message or resize paths).
+            let cur_len = last_msg_len(state);
+            if cur_len != state.cached_last_len {
+                if let Some(last_msg) = state.transcript_cache.messages.back() {
+                    if state.cached_last_start <= state.display_lines_cache.len() {
+                        let new_lines = render_one_message(last_msg, content_width);
+                        state.display_lines_cache.truncate(state.cached_last_start);
+                        state.display_lines_cache.extend(new_lines);
+                    }
+                }
+                state.cached_last_len = cur_len;
+            }
             state.transcript_cache.dirty = false;
         }
 
@@ -326,3 +349,15 @@ fn split_spans_into_lines(spans: Vec<Span<'_>>) -> Vec<Line<'_>> {
     }
     lines
 }
+
+/// Byte length of the last message's content + reasoning. Used as a cheap
+/// change detector for the streaming render branch — avoids re-rendering on
+/// spinner-only ticks where no new token arrived.
+fn last_msg_len(state: &crate::state::AppStateRest) -> usize {
+    state
+        .transcript_cache
+        .messages
+        .back()
+        .map(|m| m.content.len() + m.reasoning.len())
+        .unwrap_or(0)
+}