diff --git a/wsai/backends/claude.py b/wsai/backends/claude.py
index 3064ab3..9ab71fc 100644
--- a/wsai/backends/claude.py
+++ b/wsai/backends/claude.py
@@ -29,6 +29,13 @@ from ..interfaces import Frame, Reply, ScreenObservation
 # this identity string; the real persona/instructions go in later blocks.
 _CLAUDE_CODE_ID = "You are Claude Code, Anthropic's official CLI for Claude."
 
+# Claude occasionally returns 529 Overloaded; the anthropic SDK retries >=500
+# (and 429) with exponential backoff, but its default of 2 tries can be too few
+# to ride out a busy window. Bump it so a transient overload doesn't drop the
+# voice turn to the apology fallback. Kept modest so a *sustained* overload
+# still fails fast rather than leaving the bot silent for many seconds.
+_MAX_RETRIES = int(os.environ.get("WSAI_BRAIN_MAX_RETRIES", "4"))
+
 
 def _load_oauth_token() -> str | None:
     path = os.environ.get("CLAUDE_CREDENTIALS_PATH")
@@ -59,7 +66,7 @@ class _Auth:
         key = self._explicit_key or os.environ.get("ANTHROPIC_API_KEY")
         if key:
             if self._client is None:
-                self._client = self._anthropic.AsyncAnthropic(api_key=key)
+                self._client = self._anthropic.AsyncAnthropic(api_key=key, max_retries=_MAX_RETRIES)
             self._oauth = False
             return self._client
         tok = _load_oauth_token()
@@ -70,7 +77,7 @@ class _Auth:
             )
         if tok != self._token:
             self._token = tok
-            self._client = self._anthropic.AsyncAnthropic(auth_token=tok)
+            self._client = self._anthropic.AsyncAnthropic(auth_token=tok, max_retries=_MAX_RETRIES)
         self._oauth = True
         return self._client
 
diff --git a/wsai/dashboard.py b/wsai/dashboard.py
index b0ca831..c876492 100644
--- a/wsai/dashboard.py
+++ b/wsai/dashboard.py
@@ -304,6 +304,10 @@ class Dashboard:
         except Exception as exc:  # noqa: BLE001
             log.exception("brain failed")
             self.monitor.log("error", f"두뇌 응답 실패: {exc}")
+            blob = f"{getattr(exc, 'status_code', '')} {exc}".lower()
+            if "529" in blob or "overload" in blob:
+                # Transient server overload survived the SDK retries.
+                return "지금 서버가 잠깐 붐벼서 생각이 늦네. 잠시 뒤에 다시 말해줄래?"
             return "미안, 지금 잠깐 생각이 안 났어. 다시 말해줄래?"
         if not text:
             return "음, 뭐라고 해야 할지 모르겠어. 다시 말해줄래?"