diff --git a/docs/cpu-metrics-in-fleet.md b/docs/cpu-metrics-in-fleet.md
index 3a514dfef26..3d06c74f57c 100644
--- a/docs/cpu-metrics-in-fleet.md
+++ b/docs/cpu-metrics-in-fleet.md
@@ -378,3 +378,5 @@ metrics shown in the Fleet UI for every Agent are computed as follows from the d
   runtime (e.g. Endpoint) as well, not just Agent components managed by
   the command runtime (e.g. Beats) as we do today.
 
+* We should enhance collection to include CPU utilization for Agent monitoring components so their contributions are also
+  counted.