fix(qoder): detect errors in SSE envelope to trigger failover

- Made wrapQoderSSE async to peek at first chunk before streaming
- Parse first SSE line to check for error envelope (statusCodeValue !== 200)
- If error detected, return error Response with proper HTTP status code
- This triggers chatCore's !providerResponse.ok check and failover logic
- Fixes issue where qoder 403 quota errors were wrapped as successful streams

Before: qoder errors appeared as stream content with finish_reason: 'stop'
After: qoder errors return proper HTTP error codes, triggering provider failover

open-sse/executors/qoder.js

@@ -222,22 +222,75 @@ async function buildQoderRequestBody({ model, body, credentials, log, proxyOptio
  * and re-emit as `data: <inner>\n\n`. Errors become `data: [DONE]\n\n` plus
  * a synthetic OpenAI error chunk.
  */
-function wrapQoderSSE(response, model) {
+async function wrapQoderSSE(response, model) {
   if (!response.ok || !response.body) return response;
 
   const decoder = new TextDecoder();
   const encoder = new TextEncoder();
+
+  // Peek at first chunk to detect errors early
+  const reader = response.body.getReader();
+  const firstRead = await reader.read();
+  
+  if (firstRead.done) {
+    // Empty stream
+    return new Response("data: [DONE]\n\n", {
+      status: response.status,
+      statusText: response.statusText,
+      headers: {
+        "Content-Type": "text/event-stream",
+        "Cache-Control": "no-cache",
+      },
+    });
+  }
+
+  // Parse first line to check for error
+  const firstText = decoder.decode(firstRead.value, { stream: true });
+  const nlIndex = firstText.indexOf("\n");
+  const firstLine = nlIndex !== -1 ? firstText.slice(0, nlIndex) : firstText;
+  const trimmed = firstLine.replace(/\r$/, "").trim();
+  
+  if (trimmed.startsWith("data:")) {
+    const data = trimmed.slice(5).trimStart();
+    if (data !== "[DONE]") {
+      try {
+        const envelope = JSON.parse(data);
+        const statusVal = typeof envelope.statusCodeValue === "number" ? envelope.statusCodeValue : 200;
+        
+        if (statusVal !== 200) {
+          // Error detected - return error Response to trigger failover
+          const msg = envelope.body || `upstream status ${statusVal}`;
+          const errorResponse = new Response(
+            JSON.stringify({ 
+              error: { 
+                message: `qoder error ${statusVal}: ${truncate(msg, 500)}`,
+                type: "upstream_error",
+                code: String(statusVal)
+              } 
+            }),
+            { 
+              status: statusVal >= 400 && statusVal < 600 ? statusVal : 502,
+              headers: { "Content-Type": "application/json" }
+            }
+          );
+          reader.cancel();
+          return errorResponse;
+        }
+      } catch (e) {
+        // Not JSON, continue as normal stream
+      }
+    }
+  }
+
+  // No error detected - proceed with normal TransformStream
   let buffer = "";
   let doneEmitted = false;
 
-  // Process one already-extracted SSE line (no trailing newline). Returns
-  // false when the line indicated end-of-stream so the caller can stop
-  // forwarding any remaining chunks after [DONE].
   const processLine = (line, controller) => {
     const trimmed = line.replace(/\r$/, "").trim();
     if (!trimmed) return;
     if (!trimmed.startsWith("data:")) return;
-    if (doneEmitted) return; // never forward chunks past stream end
+    if (doneEmitted) return;
 
     const data = trimmed.slice(5).trimStart();
     if (data === "[DONE]") {
@@ -270,10 +323,6 @@ function wrapQoderSSE(response, model) {
       doneEmitted = true;
       return;
     }
-    // Inner is an OpenAI-shaped chunk. Strip any embedded newlines so the
-    // SSE frame stays a single event (a literal "\n" inside `inner` would
-    // otherwise split the frame across multiple data: lines and downstream
-    // parsers would reassemble them as separate events).
     const sanitized = inner.replace(/\r?\n/g, "");
     controller.enqueue(encoder.encode(`data: ${sanitized}\n\n`));
   };
@@ -289,13 +338,7 @@ function wrapQoderSSE(response, model) {
       }
     },
     flush(controller) {
-      // Finalize the decoder so any pending multi-byte sequence is
-      // released into `buffer` instead of being silently dropped.
       buffer += decoder.decode();
-      // Drain any trailing line that arrived without a terminating newline
-      // (e.g. upstream closed the socket immediately after the last write,
-      // or a CDN stripped the final CRLF). Without this, the chunk that
-      // carries finish_reason is silently lost.
       if (buffer.length > 0) {
         processLine(buffer, controller);
         buffer = "";
@@ -307,9 +350,45 @@ function wrapQoderSSE(response, model) {
     },
   });
 
-  const transformed = response.body.pipeThrough(transform);
+  // Create a ReadableStream that emits the first chunk + remaining chunks
-  // Build a Response with passable headers; the streaming handler reads
+  const combinedStream = new ReadableStream({
-  // `.body` as a ReadableStream regardless of Content-Type.
+    start(controller) {
+      controller.enqueue(firstRead.value);
+    },
+    async pull(controller) {
+      const { done, value } = await reader.read();
+      if (done) {
+        controller.close();
+      } else {
+        controller.enqueue(value);
+      }
+    },
+    cancel() {
+      reader.cancel();
+    }
+  });
+
+  const transformed = combinedStream.pipeThrough(transform);
+  return new Response(transformed, {
+    status: response.status,
+    statusText: response.statusText,
+    headers: {
+      "Content-Type": "text/event-stream",
+      "Cache-Control": "no-cache",
+    },
+  });
+}
+        controller.close();
+      } else {
+        controller.enqueue(value);
+      }
+    },
+    cancel() {
+      reader.cancel();
+    }
+  });
+
+  const transformed = remainingStream.pipeThrough(transform);
   return new Response(transformed, {
     status: response.status,
     statusText: response.statusText,
@@ -430,7 +509,7 @@ export class QoderExecutor extends BaseExecutor {
       return { response, url, headers, transformedBody: payload };
     }
 
-    const wrapped = wrapQoderSSE(response, `qoder/${qoderKey}`);
+    const wrapped = await wrapQoderSSE(response, `qoder/${qoderKey}`);
     return { response: wrapped, url, headers, transformedBody: payload };
   }