fix(core): retry TASK_MIDDLEWARE_ERROR under the task's retry policy

.changeset/retry-middleware-errors.md

@@ -0,0 +1,5 @@
+---
+"@trigger.dev/core": patch
+---
+
+Retry TASK_MIDDLEWARE_ERROR under the task's retry policy instead of failing the run on the first attempt. The error was already classified as retryable by shouldRetryError, but shouldLookupRetrySettings did not include it, so the retry flow fell through to fail_run. Fixes #3231.

packages/core/src/v3/errors.ts

@@ -320,7 +320,7 @@ export function sanitizeError(error: TaskRunError): TaskRunError {
     case "CUSTOM_ERROR": {
       // CUSTOM_ERROR.raw holds JSON.stringify(error) which is later parsed by
       // JSON.parse in createErrorTaskError. Naive truncation would cut mid-token
-      // and produce invalid JSON — wrap the preview in a valid JSON envelope.
+      // and produce invalid JSON — wrap the preview in a valid JSON envelope.
       const clean = error.raw.replace(/\0/g, "");
       const safeRaw =
         clean.length > MAX_MESSAGE_LENGTH
@@ -332,7 +332,7 @@ export function sanitizeError(error: TaskRunError): TaskRunError {
       };
     }
     case "INTERNAL_ERROR": {
-      // message and stackTrace are optional for INTERNAL_ERROR — preserve
+      // message and stackTrace are optional for INTERNAL_ERROR — preserve
       // `undefined` so the `error.message ?? "Internal error (CODE)"` fallback
       // in createErrorTaskError still kicks in (empty string is not nullish).
       return {
@@ -429,6 +429,9 @@ export function shouldLookupRetrySettings(error: TaskRunError): boolean {
         case "TASK_RUN_UNCAUGHT_EXCEPTION":
           return true;
 
+        case "TASK_MIDDLEWARE_ERROR":
+          return true;
+
         default:
           return false;
       }
@@ -641,7 +644,7 @@ export class ChatChunkTooLargeError extends Error {
       `chat.agent chunk${chunkType ? ` of type "${chunkType}"` : ""} is ${chunkSize} bytes, ` +
         `over the realtime stream's per-record cap of ${maxSize} bytes. ` +
         `For oversized payloads (e.g. large tool outputs), write the value to your own store and ` +
-        `emit only an id/url through the chat stream — see https://trigger.dev/docs/ai-chat/patterns/large-payloads.`
+        `emit only an id/url through the chat stream — see https://trigger.dev/docs/ai-chat/patterns/large-payloads.`
     );
     this.name = "ChatChunkTooLargeError";
   }
@@ -744,7 +747,7 @@ const prettyInternalErrors: Partial<
       href: links.docs.troubleshooting.stalledExecution,
     },
   },
-  // Link only — we deliberately do NOT set `message`, so the original
+  // Link only — we deliberately do NOT set `message`, so the original
   // error message (e.g. "read ECONNRESET") is preserved in the dashboard.
   // Common cause: an EventEmitter (node-redis, pg, etc.) emitted "error"
   // with no listener attached, which Node escalates to uncaughtException.
@@ -1152,7 +1155,7 @@ export function createTaskMetadataFailedErrorStack(
     }
 
     stack.push("\n");
-    stack.push(`  ❯ ${taskWithIssues.exportName} in ${taskWithIssues.filePath}`);
+    stack.push(`  ? ${taskWithIssues.exportName} in ${taskWithIssues.filePath}`);
 
     for (const issue of taskWithIssues.issues) {
       if (issue.path) {

packages/core/test/errors.test.ts

@@ -263,6 +263,12 @@ describe("shouldRetryError + shouldLookupRetrySettings", () => {
     expect(shouldLookupRetrySettings(err)).toBe(true);
   });
 
+    it("retries TASK_MIDDLEWARE_ERROR using the task's retry settings", () => {
+    const err = internal("TASK_MIDDLEWARE_ERROR");
+    expect(shouldRetryError(err)).toBe(true);
+    expect(shouldLookupRetrySettings(err)).toBe(true);
+  });
+
   it("still does not retry SIGKILL timeout", () => {
     expect(shouldRetryError(internal("TASK_PROCESS_SIGKILL_TIMEOUT"))).toBe(false);
   });