diff --git a/openclaw b/openclaw
new file mode 160000
index 000000000..ccac2aeec
--- /dev/null
+++ b/openclaw
@@ -0,0 +1 @@
+Subproject commit ccac2aeec2dad4bbca49707568311123edf5545a
diff --git a/src/gateway/protocol/client-info.ts b/src/gateway/protocol/client-info.ts
index 9fc39ff11..1f9a48d48 100644
--- a/src/gateway/protocol/client-info.ts
+++ b/src/gateway/protocol/client-info.ts
@@ -13,7 +13,31 @@ export const GATEWAY_CLIENT_IDS = {
   PROBE: "openclaw-probe",
 } as const;
 
-export type GatewayClientId = (typeof GATEWAY_CLIENT_IDS)[keyof typeof GATEWAY_CLIENT_IDS];
+// Legacy client IDs for backward compatibility
+// These are deprecated but still accepted to prevent breakage during upgrades
+export const LEGACY_GATEWAY_CLIENT_IDS = {
+  // Clawdbot era (pre-2026.1.29)
+  CLAWDBOT_CONTROL_UI: "clawdbot-control-ui",
+  CLAWDBOT_MACOS_APP: "clawdbot-macos",
+  CLAWDBOT_IOS_APP: "clawdbot-ios",
+  CLAWDBOT_ANDROID_APP: "clawdbot-android",
+  CLAWDBOT_PROBE: "clawdbot-probe",
+  // Moltbot era (intermediate rebrand)
+  MOLTBOT_CONTROL_UI: "moltbot-control-ui",
+  MOLTBOT_MACOS_APP: "moltbot-macos",
+  MOLTBOT_IOS_APP: "moltbot-ios",
+  MOLTBOT_ANDROID_APP: "moltbot-android",
+  MOLTBOT_PROBE: "moltbot-probe",
+} as const;
+
+export const ALL_GATEWAY_CLIENT_IDS = {
+  ...GATEWAY_CLIENT_IDS,
+  ...LEGACY_GATEWAY_CLIENT_IDS,
+} as const;
+
+export type GatewayClientId =
+  | (typeof GATEWAY_CLIENT_IDS)[keyof typeof GATEWAY_CLIENT_IDS]
+  | (typeof LEGACY_GATEWAY_CLIENT_IDS)[keyof typeof LEGACY_GATEWAY_CLIENT_IDS];
 
 // Back-compat naming (internal): these values are IDs, not display names.
 export const GATEWAY_CLIENT_NAMES = GATEWAY_CLIENT_IDS;
@@ -42,7 +66,7 @@ export type GatewayClientInfo = {
   instanceId?: string;
 };
 
-const GATEWAY_CLIENT_ID_SET = new Set<GatewayClientId>(Object.values(GATEWAY_CLIENT_IDS));
+const GATEWAY_CLIENT_ID_SET = new Set<GatewayClientId>(Object.values(ALL_GATEWAY_CLIENT_IDS));
 const GATEWAY_CLIENT_MODE_SET = new Set<GatewayClientMode>(Object.values(GATEWAY_CLIENT_MODES));
 
 export function normalizeGatewayClientId(raw?: string | null): GatewayClientId | undefined {
diff --git a/src/gateway/protocol/schema/primitives.ts b/src/gateway/protocol/schema/primitives.ts
index d43a16a1e..9e38c1dde 100644
--- a/src/gateway/protocol/schema/primitives.ts
+++ b/src/gateway/protocol/schema/primitives.ts
@@ -1,6 +1,6 @@
 import { Type } from "@sinclair/typebox";
 import { SESSION_LABEL_MAX_LENGTH } from "../../../sessions/session-label.js";
-import { GATEWAY_CLIENT_IDS, GATEWAY_CLIENT_MODES } from "../client-info.js";
+import { ALL_GATEWAY_CLIENT_IDS, GATEWAY_CLIENT_MODES } from "../client-info.js";
 
 export const NonEmptyString = Type.String({ minLength: 1 });
 export const SessionLabelString = Type.String({
@@ -9,7 +9,7 @@ export const SessionLabelString = Type.String({
 });
 
 export const GatewayClientIdSchema = Type.Union(
-  Object.values(GATEWAY_CLIENT_IDS).map((value) => Type.Literal(value)),
+  Object.values(ALL_GATEWAY_CLIENT_IDS).map((value) => Type.Literal(value)),
 );
 
 export const GatewayClientModeSchema = Type.Union(
diff --git a/src/gateway/server.ios-client-id.e2e.test.ts b/src/gateway/server.ios-client-id.e2e.test.ts
index 3c00d23fb..d4f82adf5 100644
--- a/src/gateway/server.ios-client-id.e2e.test.ts
+++ b/src/gateway/server.ios-client-id.e2e.test.ts
@@ -91,3 +91,60 @@ test("accepts openclaw-android as a valid gateway client id", async () => {
 
   ws.close();
 });
+
+test("accepts legacy clawdbot-ios as a valid gateway client id (backward compat)", async () => {
+  const ws = new WebSocket(`ws://127.0.0.1:${port}`);
+  await new Promise<void>((resolve) => ws.once("open", resolve));
+
+  const res = await connectReq(ws, { clientId: "clawdbot-ios", platform: "ios" });
+  // We don't care if auth fails here; we only care that schema validation accepts the client id.
+  // A schema rejection would close the socket before sending a response.
+  if (!res.ok) {
+    // allow unauthorized error when gateway requires auth
+    // but reject schema validation errors
+    const message = String(res.error?.message ?? "");
+    if (message.includes("invalid connect params")) {
+      throw new Error(message);
+    }
+  }
+
+  ws.close();
+});
+
+test("accepts legacy clawdbot-android as a valid gateway client id (backward compat)", async () => {
+  const ws = new WebSocket(`ws://127.0.0.1:${port}`);
+  await new Promise<void>((resolve) => ws.once("open", resolve));
+
+  const res = await connectReq(ws, { clientId: "clawdbot-android", platform: "android" });
+  // We don't care if auth fails here; we only care that schema validation accepts the client id.
+  // A schema rejection would close the socket before sending a response.
+  if (!res.ok) {
+    // allow unauthorized error when gateway requires auth
+    // but reject schema validation errors
+    const message = String(res.error?.message ?? "");
+    if (message.includes("invalid connect params")) {
+      throw new Error(message);
+    }
+  }
+
+  ws.close();
+});
+
+test("accepts legacy moltbot-macos as a valid gateway client id (backward compat)", async () => {
+  const ws = new WebSocket(`ws://127.0.0.1:${port}`);
+  await new Promise<void>((resolve) => ws.once("open", resolve));
+
+  const res = await connectReq(ws, { clientId: "moltbot-macos", platform: "macos" });
+  // We don't care if auth fails here; we only care that schema validation accepts the client id.
+  // A schema rejection would close the socket before sending a response.
+  if (!res.ok) {
+    // allow unauthorized error when gateway requires auth
+    // but reject schema validation errors
+    const message = String(res.error?.message ?? "");
+    if (message.includes("invalid connect params")) {
+      throw new Error(message);
+    }
+  }
+
+  ws.close();
+});
diff --git a/src/telegram/monitor.test.ts b/src/telegram/monitor.test.ts
index 2fc46827b..f2800bc4d 100644
--- a/src/telegram/monitor.test.ts
+++ b/src/telegram/monitor.test.ts
@@ -184,4 +184,122 @@ describe("monitorTelegramProvider (grammY)", () => {
 
     await expect(monitorTelegramProvider({ token: "tok" })).rejects.toThrow("bad token");
   });
+
+  it("retries on getUpdates timeout errors", async () => {
+    const timeoutError = Object.assign(
+      new Error("Request to 'getUpdates' timed out after 30 seconds"),
+      {
+        method: "getUpdates",
+        description: "Request timed out",
+      },
+    );
+    runSpy
+      .mockImplementationOnce(() => ({
+        task: () => Promise.reject(timeoutError),
+        stop: vi.fn(),
+      }))
+      .mockImplementationOnce(() => ({
+        task: () => Promise.resolve(),
+        stop: vi.fn(),
+      }));
+
+    await monitorTelegramProvider({ token: "tok" });
+
+    expect(computeBackoff).toHaveBeenCalled();
+    expect(sleepWithAbort).toHaveBeenCalled();
+    expect(runSpy).toHaveBeenCalledTimes(2);
+  });
+
+  it("restarts when runner exhausts retries and stops normally", async () => {
+    runSpy
+      .mockImplementationOnce(() => ({
+        task: () => Promise.resolve(), // Runner stopped normally after exhausting retries
+        stop: vi.fn(),
+      }))
+      .mockImplementationOnce(() => ({
+        task: () => Promise.resolve(),
+        stop: vi.fn(),
+      }));
+
+    await monitorTelegramProvider({ token: "tok" });
+
+    expect(computeBackoff).toHaveBeenCalled();
+    expect(sleepWithAbort).toHaveBeenCalled();
+    expect(runSpy).toHaveBeenCalledTimes(2);
+  });
+
+  it("retries on getUpdates conflict errors (409)", async () => {
+    const conflictError = Object.assign(new Error("Conflict: terminated by other getUpdates"), {
+      error_code: 409,
+      description: "Conflict: terminated by other getUpdates request",
+      method: "getUpdates",
+    });
+    runSpy
+      .mockImplementationOnce(() => ({
+        task: () => Promise.reject(conflictError),
+        stop: vi.fn(),
+      }))
+      .mockImplementationOnce(() => ({
+        task: () => Promise.resolve(),
+        stop: vi.fn(),
+      }));
+
+    await monitorTelegramProvider({ token: "tok" });
+
+    expect(computeBackoff).toHaveBeenCalled();
+    expect(sleepWithAbort).toHaveBeenCalled();
+    expect(runSpy).toHaveBeenCalledTimes(2);
+  });
+
+  it("respects abort signal during retry backoff", async () => {
+    const abortController = new AbortController();
+    const networkError = Object.assign(new Error("timeout"), { code: "ETIMEDOUT" });
+
+    // Mock sleepWithAbort to abort mid-sleep
+    sleepWithAbort.mockImplementationOnce(async () => {
+      abortController.abort();
+      throw new Error("Aborted");
+    });
+
+    runSpy.mockImplementationOnce(() => ({
+      task: () => Promise.reject(networkError),
+      stop: vi.fn(),
+    }));
+
+    await monitorTelegramProvider({ token: "tok", abortSignal: abortController.signal });
+
+    expect(runSpy).toHaveBeenCalledTimes(1);
+    expect(computeBackoff).toHaveBeenCalled();
+    expect(sleepWithAbort).toHaveBeenCalled();
+  });
+
+  it("uses exponential backoff for consecutive failures", async () => {
+    computeBackoff.mockReturnValueOnce(2000).mockReturnValueOnce(3600).mockReturnValueOnce(6480);
+
+    runSpy
+      .mockImplementationOnce(() => ({
+        task: () => Promise.reject(Object.assign(new Error("timeout"), { code: "ETIMEDOUT" })),
+        stop: vi.fn(),
+      }))
+      .mockImplementationOnce(() => ({
+        task: () => Promise.reject(Object.assign(new Error("timeout"), { code: "ETIMEDOUT" })),
+        stop: vi.fn(),
+      }))
+      .mockImplementationOnce(() => ({
+        task: () => Promise.reject(Object.assign(new Error("timeout"), { code: "ETIMEDOUT" })),
+        stop: vi.fn(),
+      }))
+      .mockImplementationOnce(() => ({
+        task: () => Promise.resolve(),
+        stop: vi.fn(),
+      }));
+
+    await monitorTelegramProvider({ token: "tok" });
+
+    // Verify backoff was called with increasing attempt numbers
+    expect(computeBackoff).toHaveBeenCalledTimes(4); // 3 failures + 1 normal stop
+    expect(computeBackoff).toHaveBeenCalledWith(expect.anything(), 1);
+    expect(computeBackoff).toHaveBeenCalledWith(expect.anything(), 2);
+    expect(computeBackoff).toHaveBeenCalledWith(expect.anything(), 3);
+  });
 });
diff --git a/src/telegram/monitor.ts b/src/telegram/monitor.ts
index 2709b591b..334c12310 100644
--- a/src/telegram/monitor.ts
+++ b/src/telegram/monitor.ts
@@ -74,6 +74,22 @@ const isGetUpdatesConflict = (err: unknown) => {
   return haystack.includes("getupdates");
 };
 
+const isGetUpdatesTimeout = (err: unknown) => {
+  if (!err || typeof err !== "object") return false;
+  const typed = err as {
+    description?: string;
+    method?: string;
+    message?: string;
+    error_description?: string;
+  };
+  const haystack = [typed.method, typed.description, typed.message, typed.error_description]
+    .filter((value): value is string => typeof value === "string")
+    .join(" ")
+    .toLowerCase();
+
+  return haystack.includes("getupdates") && haystack.includes("timeout");
+};
+
 const NETWORK_ERROR_SNIPPETS = [
   "fetch failed",
   "network",
@@ -168,20 +184,41 @@ export async function monitorTelegramProvider(opts: MonitorTelegramOpts = {}) {
     try {
       // runner.task() returns a promise that resolves when the runner stops
       await runner.task();
-      return;
+      // Runner stopped normally (aborted or exhausted retries).
+      // If not aborted, this is a transient failure - retry with backoff.
+      if (opts.abortSignal?.aborted) {
+        return;
+      }
+      // grammY runner exhausted its maxRetryTime - treat as recoverable network error
+      restartAttempts += 1;
+      const delayMs = computeBackoff(TELEGRAM_POLL_RESTART_POLICY, restartAttempts);
+      (opts.runtime?.error ?? console.error)(
+        `Telegram polling stopped after retry timeout; restarting in ${formatDurationMs(delayMs)}.`,
+      );
+      try {
+        await sleepWithAbort(delayMs, opts.abortSignal);
+      } catch (sleepErr) {
+        if (opts.abortSignal?.aborted) return;
+        throw sleepErr;
+      }
     } catch (err) {
       if (opts.abortSignal?.aborted) {
         throw err;
       }
       const isConflict = isGetUpdatesConflict(err);
+      const isTimeout = isGetUpdatesTimeout(err);
       const isRecoverable = isRecoverableTelegramNetworkError(err, { context: "polling" });
       const isNetworkError = isNetworkRelatedError(err);
-      if (!isConflict && !isRecoverable && !isNetworkError) {
+      if (!isConflict && !isTimeout && !isRecoverable && !isNetworkError) {
         throw err;
       }
       restartAttempts += 1;
       const delayMs = computeBackoff(TELEGRAM_POLL_RESTART_POLICY, restartAttempts);
-      const reason = isConflict ? "getUpdates conflict" : "network error";
+      const reason = isConflict
+        ? "getUpdates conflict"
+        : isTimeout
+          ? "getUpdates timeout"
+          : "network error";
       const errMsg = formatErrorMessage(err);
       (opts.runtime?.error ?? console.error)(
         `Telegram ${reason}: ${errMsg}; retrying in ${formatDurationMs(delayMs)}.`,