diff --git a/.openclaw-sync/source.json b/.openclaw-sync/source.json
index d54429966..ebe56034a 100644
--- a/.openclaw-sync/source.json
+++ b/.openclaw-sync/source.json
@@ -1,5 +1,5 @@
 {
   "repository": "openclaw/openclaw",
-  "sha": "2805bbd3d7a95dbf31dd9e69a1d3aed660c1bb1f",
-  "syncedAt": "2026-05-03T17:23:51.233Z"
+  "sha": "68918ab36a5660def8ce04e113161187846baf34",
+  "syncedAt": "2026-05-03T17:29:38.534Z"
 }
diff --git a/docs/tools/loop-detection.md b/docs/tools/loop-detection.md
index 7e3a88726..61a5d644d 100644
--- a/docs/tools/loop-detection.md
+++ b/docs/tools/loop-detection.md
@@ -78,7 +78,7 @@ When a run id is available, recent tool-call history is evaluated only within th
 
 ## Recommended setup
 
-- Start with `enabled: true`, defaults unchanged.
+- For smaller models, start with `enabled: true`, defaults unchanged. Flagship models rarely need loop detection and can leave it disabled.
 - Keep thresholds ordered as `warningThreshold < criticalThreshold < globalCircuitBreakerThreshold`.
 - If false positives occur:
   - raise `warningThreshold` and/or `criticalThreshold`