diff --git a/package.json b/package.json
index 236dae5..483bab7 100644
--- a/package.json
+++ b/package.json
@@ -74,7 +74,7 @@
     "limiter": "^3.0.0",
     "marked": "^16.1.1",
     "node-fetch": "^3.3.2",
-    "p-queue": "^8.1.0",
+    "p-queue": "^9.0.0",
     "puppeteer": "^24.10.1",
     "sass": "^1.89.2",
     "strict-csp": "^1.1.1",
diff --git a/pnpm-lock.yaml b/pnpm-lock.yaml
index 322c74d..8236481 100644
--- a/pnpm-lock.yaml
+++ b/pnpm-lock.yaml
@@ -78,8 +78,8 @@ importers:
         specifier: ^3.3.2
         version: 3.3.2
       p-queue:
-        specifier: ^8.1.0
-        version: 8.1.1
+        specifier: ^9.0.0
+        version: 9.0.0
       puppeteer:
         specifier: ^24.10.1
         version: 24.22.3(typescript@5.9.3)
@@ -5288,13 +5288,13 @@ packages:
     resolution: {integrity: sha512-VkndIv2fIB99swvQoA65bm+fsmt6UNdGeIB0oxBs+WhAhdh08QA04JXpI7rbB9r08/nkbysKoya9rtDERYOYMA==}
     engines: {node: '>=18'}
 
-  p-queue@8.1.1:
-    resolution: {integrity: sha512-aNZ+VfjobsWryoiPnEApGGmf5WmNsCo9xu8dfaYamG5qaLP7ClhLN6NgsFe6SwJ2UbLEBK5dv9x8Mn5+RVhMWQ==}
-    engines: {node: '>=18'}
+  p-queue@9.0.0:
+    resolution: {integrity: sha512-KO1RyxstL9g1mK76530TExamZC/S2Glm080Nx8PE5sTd7nlduDQsAfEl4uXX+qZjLiwvDauvzXavufy3+rJ9zQ==}
+    engines: {node: '>=20'}
 
-  p-timeout@6.1.4:
-    resolution: {integrity: sha512-MyIV3ZA/PmyBN/ud8vV9XzwTrNtR4jFrObymZYnZqMmW0zA8Z17vnT0rBgFE/TlohB+YCHqXMgZzb3Csp49vqg==}
-    engines: {node: '>=14.16'}
+  p-timeout@7.0.0:
+    resolution: {integrity: sha512-DhZ7ydOE3JXtXzDf2wz/KEamkKAD7Il5So09I2tOz4i+9pLcdghDKKmODkkoHKJ0TyczmhcHNxyTeTrsENT81A==}
+    engines: {node: '>=20'}
 
   pac-proxy-agent@7.2.0:
     resolution: {integrity: sha512-TEB8ESquiLMc0lV8vcd5Ql/JAKAoyzHFXaStwjkzpOpC5Yv+pIzLfHvjTSdf3vpa2bMiUQrg9i6276yn8666aA==}
@@ -12821,12 +12821,12 @@ snapshots:
 
   p-map@7.0.3: {}
 
-  p-queue@8.1.1:
+  p-queue@9.0.0:
     dependencies:
       eventemitter3: 5.0.1
-      p-timeout: 6.1.4
+      p-timeout: 7.0.0
 
-  p-timeout@6.1.4: {}
+  p-timeout@7.0.0: {}
 
   pac-proxy-agent@7.2.0:
     dependencies:
diff --git a/runner/orchestration/gateways/local_gateway.ts b/runner/orchestration/gateways/local_gateway.ts
index 75cd970..066c159 100644
--- a/runner/orchestration/gateways/local_gateway.ts
+++ b/runner/orchestration/gateways/local_gateway.ts
@@ -84,7 +84,6 @@ export class LocalGateway implements Gateway<LocalEnvironment> {
             reject(err);
           });
         }),
-      {throwOnTimeout: true},
     );
   }
 
diff --git a/runner/orchestration/generate.ts b/runner/orchestration/generate.ts
index 95b12c4..16bcfd2 100644
--- a/runner/orchestration/generate.ts
+++ b/runner/orchestration/generate.ts
@@ -152,60 +152,57 @@ export async function generateCodeAndAssess(options: {
 
     for (const rootPromptDef of promptsToProcess) {
       allTasks.push(
-        appConcurrencyQueue.add(
-          async () => {
-            const evalID = await env.gateway.initializeEval();
-            let results: AssessmentResult[] | undefined;
-
-            try {
-              results = await callWithTimeout(
-                `Evaluation of ${rootPromptDef.name}`,
-                async abortSignal =>
-                  startEvaluationTask(
-                    evalID,
-                    env,
-                    env.gateway,
-                    ratingLlm,
-                    options.model,
-                    rootPromptDef,
-                    options.localMode,
-                    options.skipScreenshots,
-                    options.outputDirectory,
-                    options.ragEndpoint,
-                    abortSignal,
-                    options.skipAxeTesting,
-                    !!options.enableUserJourneyTesting,
-                    !!options.enableAutoCsp,
-                    workerConcurrencyQueue,
-                    progress,
-                    options.autoraterModel || DEFAULT_AUTORATER_MODEL_NAME,
-                    options.a11yRepairAttempts ?? 0,
-                  ),
-                // 10min max per app evaluation.  We just want to make sure it never gets stuck.
-                10,
-              );
-              return results;
-            } catch (e: unknown) {
-              failedPrompts.push({
-                promptName: rootPromptDef.name,
-                error: `${e}`,
-                stack: e instanceof Error ? e.stack : undefined,
-              });
-
-              let details = `Error: ${e}`;
-              if (e instanceof Error && e.stack) {
-                details += `\nStack: ${e.stack}`;
-              }
-
-              progress.log(rootPromptDef, 'error', 'Failed to evaluate code', details);
-              return [] satisfies AssessmentResult[];
-            } finally {
-              progress.evalFinished(rootPromptDef, results || []);
-              await env.gateway.finalizeEval(evalID);
+        appConcurrencyQueue.add(async () => {
+          const evalID = await env.gateway.initializeEval();
+          let results: AssessmentResult[] | undefined;
+
+          try {
+            results = await callWithTimeout(
+              `Evaluation of ${rootPromptDef.name}`,
+              async abortSignal =>
+                startEvaluationTask(
+                  evalID,
+                  env,
+                  env.gateway,
+                  ratingLlm,
+                  options.model,
+                  rootPromptDef,
+                  options.localMode,
+                  options.skipScreenshots,
+                  options.outputDirectory,
+                  options.ragEndpoint,
+                  abortSignal,
+                  options.skipAxeTesting,
+                  !!options.enableUserJourneyTesting,
+                  !!options.enableAutoCsp,
+                  workerConcurrencyQueue,
+                  progress,
+                  options.autoraterModel || DEFAULT_AUTORATER_MODEL_NAME,
+                  options.a11yRepairAttempts ?? 0,
+                ),
+              // 10min max per app evaluation.  We just want to make sure it never gets stuck.
+              10,
+            );
+            return results;
+          } catch (e: unknown) {
+            failedPrompts.push({
+              promptName: rootPromptDef.name,
+              error: `${e}`,
+              stack: e instanceof Error ? e.stack : undefined,
+            });
+
+            let details = `Error: ${e}`;
+            if (e instanceof Error && e.stack) {
+              details += `\nStack: ${e.stack}`;
             }
-          },
-          {throwOnTimeout: true},
-        ),
+
+            progress.log(rootPromptDef, 'error', 'Failed to evaluate code', details);
+            return [] satisfies AssessmentResult[];
+          } finally {
+            progress.evalFinished(rootPromptDef, results || []);
+            await env.gateway.finalizeEval(evalID);
+          }
+        }),
       );
     }
 
diff --git a/runner/orchestration/serve-testing-worker.ts b/runner/orchestration/serve-testing-worker.ts
index 5406b13..83835e6 100644
--- a/runner/orchestration/serve-testing-worker.ts
+++ b/runner/orchestration/serve-testing-worker.ts
@@ -73,7 +73,6 @@ export async function serveAndTestApp(
               reject(err);
             });
           }),
-        {throwOnTimeout: true},
       );
     },
   );