diff --git a/cmd/estargz/README.md b/cmd/estargz/README.md
index 4459e00..3e916d7 100644
--- a/cmd/estargz/README.md
+++ b/cmd/estargz/README.md
@@ -10,7 +10,11 @@ You can use this with the containerd stargz-snapshotter to pull image contents o
 Follow [the Kubernetes setup guide](https://github.com/containerd/stargz-snapshotter#quick-start-with-kubernetes) then create a Pod that pulls an optimized image:
 
 ```
+kind create cluster --name stargz-demo --image ghcr.io/stargz-containers/estargz-kind-node:0.7.0
 kubectl run demo --image=estargz.kontain.me/nginx
 ```
 
 The first request to pull the image will cause the image to be optimized and cached, and subsequent pulls will be faster.
+Manifests and layers are cached for one day and after that they'll be rebuilt the next time they're requested.
+
+Optimized images can't be requested by the digest of the unoptimized image, because the optimized image digest won't match the original image's digest.
diff --git a/cmd/flatten/README.md b/cmd/flatten/README.md
index 70a105a..d246b50 100644
--- a/cmd/flatten/README.md
+++ b/cmd/flatten/README.md
@@ -21,3 +21,5 @@ Or by tag:
 ```
 docker pull flatten.kontain.me/busybox:musl
 ```
+
+Flattened images can't be requested by the digest of the unflattened image, because the flattened image digest won't match the original image's digest.