diff --git a/CHANGELOG.md b/CHANGELOG.md
index a974f0f57..c4073193b 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -60,6 +60,7 @@ Status: unreleased.
 - **BREAKING:** Gateway auth mode "none" is removed; gateway now requires token/password (Tailscale Serve identity still allowed).
 
 ### Fixes
+- macOS: auto-scroll to bottom when sending a new message while scrolled up. (#2471) Thanks @kennyklee.
 - Gateway: suppress AbortError and transient network errors in unhandled rejections. (#2451) Thanks @Glucksberg.
 - TTS: keep /tts status replies on text-only commands and avoid duplicate block-stream audio. (#2451) Thanks @Glucksberg.
 - Security: pin npm overrides to keep tar@7.5.4 for install toolchains.
diff --git a/apps/shared/ClawdbotKit/Sources/ClawdbotChatUI/ChatView.swift b/apps/shared/ClawdbotKit/Sources/ClawdbotChatUI/ChatView.swift
index 44399a3e6..99b32779f 100644
--- a/apps/shared/ClawdbotKit/Sources/ClawdbotChatUI/ChatView.swift
+++ b/apps/shared/ClawdbotKit/Sources/ClawdbotChatUI/ChatView.swift
@@ -13,6 +13,7 @@ public struct ClawdbotChatView: View {
     @State private var showSessions = false
     @State private var hasPerformedInitialScroll = false
     @State private var isPinnedToBottom = true
+    @State private var lastUserMessageID: UUID?
     private let showsSessionSwitcher: Bool
     private let style: Style
     private let markdownVariant: ChatMarkdownVariant
@@ -132,8 +133,28 @@ public struct ClawdbotChatView: View {
             self.hasPerformedInitialScroll = false
             self.isPinnedToBottom = true
         }
+        .onChange(of: self.viewModel.isSending) { _, isSending in
+            // Scroll to bottom when user sends a message, even if scrolled up.
+            guard isSending, self.hasPerformedInitialScroll else { return }
+            self.isPinnedToBottom = true
+            withAnimation(.snappy(duration: 0.22)) {
+                self.scrollPosition = self.scrollerBottomID
+            }
+        }
         .onChange(of: self.viewModel.messages.count) { _, _ in
-            guard self.hasPerformedInitialScroll, self.isPinnedToBottom else { return }
+            guard self.hasPerformedInitialScroll else { return }
+            if let lastMessage = self.viewModel.messages.last,
+               lastMessage.role.lowercased() == "user",
+               lastMessage.id != self.lastUserMessageID {
+                self.lastUserMessageID = lastMessage.id
+                self.isPinnedToBottom = true
+                withAnimation(.snappy(duration: 0.22)) {
+                    self.scrollPosition = self.scrollerBottomID
+                }
+                return
+            }
+
+            guard self.isPinnedToBottom else { return }
             withAnimation(.snappy(duration: 0.22)) {
                 self.scrollPosition = self.scrollerBottomID
             }
diff --git a/docs/docs.json b/docs/docs.json
index 01a338a18..f4f6258a9 100644
--- a/docs/docs.json
+++ b/docs/docs.json
@@ -956,6 +956,7 @@
           "gateway/doctor",
           "gateway/logging",
           "gateway/security",
+          "security/formal-verification",
           "gateway/sandbox-vs-tool-policy-vs-elevated",
           "gateway/sandboxing",
           "gateway/troubleshooting",
diff --git a/docs/gateway/security.md b/docs/gateway/security.md
index ca532aae0..f83ea3eb6 100644
--- a/docs/gateway/security.md
+++ b/docs/gateway/security.md
@@ -7,6 +7,8 @@ read_when:
 
 ## Quick check: `clawdbot security audit`
 
+See also: [Formal Verification (Security Models)](/security/formal-verification/)
+
 Run this regularly (especially after changing config or exposing network surfaces):
 
 ```bash
diff --git a/docs/security/formal-verification.md b/docs/security/formal-verification.md
index e2704e7b3..6028b8291 100644
--- a/docs/security/formal-verification.md
+++ b/docs/security/formal-verification.md
@@ -1,7 +1,7 @@
 ---
 title: Formal Verification (Security Models)
 summary: Machine-checked security models for Clawdbot’s highest-risk paths.
-status: draft
+permalink: /security/formal-verification/
 ---
 
 # Formal Verification (Security Models)