diff --git a/CHANGELOG.md b/CHANGELOG.md
index 6fa98109a..ae5894203 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -24,6 +24,7 @@ Docs: https://docs.openclaw.ai
 - Agents/Failover reason classification: avoid false rate-limit classification from incidental `tpm` substrings by matching TPM as a standalone token/phrase and keeping auth-context errors on the auth path. Landed from contributor PR #31007 by @HOYALIM. Thanks @HOYALIM.
 - Slack/Announce target account routing: enable session-backed announce-target lookup for Slack so multi-account announces resolve the correct `accountId` instead of defaulting to bot-token context. Landed from contributor PR #31028 by @taw0002. Thanks @taw0002.
 - Tools/Edit workspace boundary errors: preserve the real `Path escapes workspace root` failure path instead of surfacing a misleading access/file-not-found error when editing outside workspace roots. Landed from contributor PR #31015 by @haosenwang1018. Thanks @haosenwang1018.
+- Auto-reply/NO_REPLY: strip `NO_REPLY` token from mixed-content messages instead of leaking raw control text to end users. Landed from contributor PR #31080 by @scoootscooob. Thanks @scoootscooob.
 - Sandbox/mkdirp boundary checks: allow directory-safe boundary validation for existing in-boundary subdirectories, preventing false `cannot create directories` failures in sandbox write mode. (#30610) Thanks @glitch418x.
 - Android/Voice screen TTS: stream assistant speech via ElevenLabs WebSocket in Talk Mode, stop cleanly on speaker mute/barge-in, and ignore stale out-of-order stream events. (#29521) Thanks @gregmousseau.
 - Web UI/Cron: include configured agent model defaults/fallbacks in cron model suggestions so scheduled-job model autocomplete reflects configured models. (#29709) Thanks @Sid-Qin.
diff --git a/src/auto-reply/reply/normalize-reply.ts b/src/auto-reply/reply/normalize-reply.ts
index 0436b1a1d..9aafb66bd 100644
--- a/src/auto-reply/reply/normalize-reply.ts
+++ b/src/auto-reply/reply/normalize-reply.ts
@@ -1,6 +1,11 @@
 import { sanitizeUserFacingText } from "../../agents/pi-embedded-helpers.js";
 import { stripHeartbeatToken } from "../heartbeat.js";
-import { HEARTBEAT_TOKEN, isSilentReplyText, SILENT_REPLY_TOKEN } from "../tokens.js";
+import {
+  HEARTBEAT_TOKEN,
+  isSilentReplyText,
+  SILENT_REPLY_TOKEN,
+  stripSilentToken,
+} from "../tokens.js";
 import type { ReplyPayload } from "../types.js";
 import { hasLineDirectives, parseLineDirectives } from "./line-directives.js";
 import {
@@ -43,6 +48,16 @@ export function normalizeReplyPayload(
     }
     text = "";
   }
+  // Strip NO_REPLY from mixed-content messages (e.g. "😄 NO_REPLY") so the
+  // token never leaks to end users.  If stripping leaves nothing, treat it as
+  // silent just like the exact-match path above.  (#30916, #30955)
+  if (text && text.includes(silentToken) && !isSilentReplyText(text, silentToken)) {
+    text = stripSilentToken(text, silentToken);
+    if (!text && !hasMedia && !hasChannelData) {
+      opts.onSkip?.("silent");
+      return null;
+    }
+  }
   if (text && !trimmed) {
     // Keep empty text when media exists so media-only replies still send.
     text = "";
diff --git a/src/auto-reply/reply/reply-utils.test.ts b/src/auto-reply/reply/reply-utils.test.ts
index fff937187..00c5f02e9 100644
--- a/src/auto-reply/reply/reply-utils.test.ts
+++ b/src/auto-reply/reply/reply-utils.test.ts
@@ -108,6 +108,48 @@ describe("normalizeReplyPayload", () => {
       expect(reasons, testCase.name).toEqual([testCase.reason]);
     }
   });
+
+  it("strips NO_REPLY from mixed emoji message (#30916)", () => {
+    const result = normalizeReplyPayload({ text: "😄 NO_REPLY" });
+    expect(result).not.toBeNull();
+    expect(result!.text).toContain("😄");
+    expect(result!.text).not.toContain("NO_REPLY");
+  });
+
+  it("strips NO_REPLY appended after substantive text (#30916)", () => {
+    const result = normalizeReplyPayload({
+      text: "File's there. Not urgent.\n\nNO_REPLY",
+    });
+    expect(result).not.toBeNull();
+    expect(result!.text).toContain("File's there");
+    expect(result!.text).not.toContain("NO_REPLY");
+  });
+
+  it("keeps NO_REPLY when used as leading substantive text", () => {
+    const result = normalizeReplyPayload({ text: "NO_REPLY -- nope" });
+    expect(result).not.toBeNull();
+    expect(result!.text).toBe("NO_REPLY -- nope");
+  });
+
+  it("suppresses message when stripping NO_REPLY leaves nothing", () => {
+    const reasons: string[] = [];
+    const result = normalizeReplyPayload(
+      { text: "  NO_REPLY  " },
+      { onSkip: (reason) => reasons.push(reason) },
+    );
+    expect(result).toBeNull();
+    expect(reasons).toEqual(["silent"]);
+  });
+
+  it("strips NO_REPLY but keeps media payload", () => {
+    const result = normalizeReplyPayload({
+      text: "NO_REPLY",
+      mediaUrl: "https://example.com/img.png",
+    });
+    expect(result).not.toBeNull();
+    expect(result!.text).toBe("");
+    expect(result!.mediaUrl).toBe("https://example.com/img.png");
+  });
 });
 
 describe("typing controller", () => {
diff --git a/src/auto-reply/tokens.test.ts b/src/auto-reply/tokens.test.ts
index 262932f82..6dc51d1b7 100644
--- a/src/auto-reply/tokens.test.ts
+++ b/src/auto-reply/tokens.test.ts
@@ -1,5 +1,5 @@
 import { describe, it, expect } from "vitest";
-import { isSilentReplyPrefixText, isSilentReplyText } from "./tokens.js";
+import { isSilentReplyPrefixText, isSilentReplyText, stripSilentToken } from "./tokens.js";
 
 describe("isSilentReplyText", () => {
   it("returns true for exact token", () => {
@@ -36,6 +36,37 @@ describe("isSilentReplyText", () => {
   });
 });
 
+describe("stripSilentToken", () => {
+  it("strips token from end of text", () => {
+    expect(stripSilentToken("Done.\n\nNO_REPLY")).toBe("Done.");
+  });
+
+  it("does not strip token from start of text", () => {
+    expect(stripSilentToken("NO_REPLY 👍")).toBe("NO_REPLY 👍");
+  });
+
+  it("strips token with emoji (#30916)", () => {
+    expect(stripSilentToken("😄 NO_REPLY")).toBe("😄");
+  });
+
+  it("does not strip embedded token suffix without whitespace delimiter", () => {
+    expect(stripSilentToken("interject.NO_REPLY")).toBe("interject.NO_REPLY");
+  });
+
+  it("strips only trailing occurrence", () => {
+    expect(stripSilentToken("NO_REPLY ok NO_REPLY")).toBe("NO_REPLY ok");
+  });
+
+  it("returns empty string when only token remains", () => {
+    expect(stripSilentToken("NO_REPLY")).toBe("");
+    expect(stripSilentToken("  NO_REPLY  ")).toBe("");
+  });
+
+  it("works with custom token", () => {
+    expect(stripSilentToken("done HEARTBEAT_OK", "HEARTBEAT_OK")).toBe("done");
+  });
+});
+
 describe("isSilentReplyPrefixText", () => {
   it("matches uppercase underscore prefixes", () => {
     expect(isSilentReplyPrefixText("NO_")).toBe(true);
diff --git a/src/auto-reply/tokens.ts b/src/auto-reply/tokens.ts
index 0aeda237e..9be470d64 100644
--- a/src/auto-reply/tokens.ts
+++ b/src/auto-reply/tokens.ts
@@ -17,6 +17,16 @@ export function isSilentReplyText(
   return new RegExp(`^\\s*${escaped}\\s*$`).test(text);
 }
 
+/**
+ * Strip a trailing silent reply token from mixed-content text.
+ * Returns the remaining text with the token removed (trimmed).
+ * If the result is empty, the entire message should be treated as silent.
+ */
+export function stripSilentToken(text: string, token: string = SILENT_REPLY_TOKEN): string {
+  const escaped = escapeRegExp(token);
+  return text.replace(new RegExp(`(?:^|\\s+)${escaped}\\s*$`), "").trim();
+}
+
 export function isSilentReplyPrefixText(
   text: string | undefined,
   token: string = SILENT_REPLY_TOKEN,