fix(desktop): backoff on errors to stop CPU abuse + shrink settings window (#633)

Three bugs combine to burn 130% CPU when sidecar auth fails:

1. RefreshScheduler resets backoff multiplier to 1 (fastest) on error,
   causing failed endpoints to poll at base interval instead of backing off.
   Fix: exponential backoff on errors, same as unchanged-data path.

2. classify-event batch system ignores 401 (auth failure) — only pauses
   on 429/5xx. Hundreds of classify calls fire every 2s, each wasted.
   Fix: pause 120s on 401, matching the 429/5xx pattern.

3. Fetch patch retries every 401 (refresh token + retry), doubling all
   requests to the sidecar even when token refresh consistently fails.
   Fix: 60s cooldown after a retry-401 still returns 401.

Also shrinks settings window from 760→600px (min 620→480) to reduce
the empty whitespace below content on all tabs.

src-tauri/Cargo.lock

@@ -5423,7 +5423,7 @@ dependencies = [
 
 [[package]]
 name = "world-monitor"
-version = "2.5.19"
+version = "2.5.21"
 dependencies = [
  "getrandom 0.2.17",
  "keyring",

src-tauri/src/main.rs

@@ -563,8 +563,8 @@ fn open_settings_window(app: &AppHandle) -> Result<(), String> {
 
     let _settings_window = WebviewWindowBuilder::new(app, "settings", WebviewUrl::App("settings.html".into()))
         .title("World Monitor Settings")
-        .inner_size(980.0, 760.0)
-        .min_inner_size(820.0, 620.0)
+        .inner_size(980.0, 600.0)
+        .min_inner_size(820.0, 480.0)
         .resizable(true)
         .background_color(tauri::webview::Color(26, 28, 30, 255))
         .build()

src/app/refresh-scheduler.ts

@@ -85,7 +85,7 @@ export class RefreshScheduler implements AppModule {
         }
       } catch (e) {
         console.error(`[App] Refresh ${name} failed:`, e);
-        currentMultiplier = 1;
+        currentMultiplier = Math.min(currentMultiplier * 2, MAX_BACKOFF_MULTIPLIER);
       } finally {
         this.ctx.inFlight.delete(name);
         scheduleNext(computeDelay(intervalMs * currentMultiplier, false));

src/services/runtime.ts

@@ -257,6 +257,7 @@ export function installRuntimeFetchPatch(): void {
   const nativeFetch = window.fetch.bind(window);
   let localApiToken: string | null = null;
   let tokenFetchedAt = 0;
+  let authRetryCooldownUntil = 0; // suppress 401 retries after consecutive failures
 
   window.fetch = async (input: RequestInfo | URL, init?: RequestInit): Promise<Response> => {
     const target = getApiTargetFromRequestInput(input);
@@ -333,7 +334,8 @@ export function installRuntimeFetchPatch(): void {
       if (debug) console.log(`[fetch] ${target} → ${response.status} (${Math.round(performance.now() - t0)}ms)`);
 
       // Token may be stale after a sidecar restart — refresh and retry once.
-      if (response.status === 401 && localApiToken) {
+      // Skip retry if we recently failed (avoid doubling every request during auth outages).
+      if (response.status === 401 && localApiToken && Date.now() > authRetryCooldownUntil) {
         if (debug) console.log(`[fetch] 401 from sidecar, refreshing token and retrying`);
         try {
           const { tryInvokeTauri } = await import('@/services/tauri-bridge');
@@ -348,6 +350,12 @@ export function installRuntimeFetchPatch(): void {
           retryHeaders.set('Authorization', `Bearer ${localApiToken}`);
           response = await fetchLocalWithStartupRetry(nativeFetch, localUrl, { ...init, headers: retryHeaders });
           if (debug) console.log(`[fetch] retry ${target} → ${response.status}`);
+          if (response.status === 401) {
+            authRetryCooldownUntil = Date.now() + 60_000;
+            if (debug) console.log(`[fetch] auth retry failed, suppressing retries for 60s`);
+          } else {
+            authRetryCooldownUntil = 0;
+          }
         }
       }
 

src/services/threat-classifier.ts

@@ -462,9 +462,9 @@ function flushBatch(): void {
           });
           job.resolve(toThreat(resp));
         } catch (err) {
-          if (err instanceof ApiError && (err.statusCode === 429 || err.statusCode >= 500)) {
+          if (err instanceof ApiError && (err.statusCode === 401 || err.statusCode === 429 || err.statusCode >= 500)) {
             batchPaused = true;
-            const delay = err.statusCode === 429 ? 60_000 : 30_000;
+            const delay = err.statusCode === 401 ? 120_000 : err.statusCode === 429 ? 60_000 : 30_000;
             console.warn(`[Classify] ${err.statusCode} — pausing AI classification for ${delay / 1000}s`);
             const remaining = batch.slice(i + 1);
             // Failed job: increment attempts, requeue if under limit