diff --git a/.changeset/cap-background-command-output.md b/.changeset/cap-background-command-output.md
index 09ed4bf24..9832f9c3d 100644
--- a/.changeset/cap-background-command-output.md
+++ b/.changeset/cap-background-command-output.md
@@ -1,5 +1,5 @@
 ---
-"@moonshot-ai/agent-core": minor
+"@moonshot-ai/agent-core": patch
 ---
 
 Apply the 16 MiB output cap to background shell commands too, so a runaway background command can no longer fill the disk or crash the process; it is now terminated with the same guidance to redirect large output to a file.
diff --git a/.changeset/print-drain-subagents.md b/.changeset/print-drain-subagents.md
index 6d1e56480..dde3d69ec 100644
--- a/.changeset/print-drain-subagents.md
+++ b/.changeset/print-drain-subagents.md
@@ -1,5 +1,5 @@
 ---
-"@moonshot-ai/kimi-code": minor
+"@moonshot-ai/kimi-code": patch
 ---
 
 Wait for background subagents to finish and respond to their results before exiting in `kimi -p`, instead of ending the turn early.
diff --git a/.changeset/server-bypass-auth-keep-alive.md b/.changeset/server-bypass-auth-keep-alive.md
index 88e9e22dd..0390eabef 100644
--- a/.changeset/server-bypass-auth-keep-alive.md
+++ b/.changeset/server-bypass-auth-keep-alive.md
@@ -1,5 +1,5 @@
 ---
-"@moonshot-ai/kimi-code": minor
+"@moonshot-ai/kimi-code": patch
 ---
 
 Add `--dangerous-bypass-auth` and `--keep-alive` flags to `kimi server run`, so the server can run without a token on trusted networks and stay alive past the idle timeout.