fix(llm): surface openai stream errors

Author: MouseWW

RELEASE_NOTES.md

@@ -1,16 +1,16 @@
-# Anything Analyzer v3.6.16
+# Anything Analyzer v3.6.17
 
 ## 修复
 
-- **Anthropic/MiniMax 流式末尾事件不再丢失** — 修复 Anthropic SSE 响应没有尾随换行时最后一行 `data:` 未被解析，导致末尾文本块被忽略的问题
-  - 在流结束后补处理剩余 buffer，与 OpenAI/Responses 流式解析保持一致
-  - 增加无尾随换行 SSE 回归测试，确保最后文本块会进入增量回调和最终内容
+- **OpenAI Chat 流式错误不再被吞掉** — 修复 Chat Completions SSE 返回 `error` payload 时被当成空成功响应的问题
+  - 将流式 JSON 容错解析与 API 错误判断拆开，保留 malformed chunk 跳过逻辑
+  - 增加错误 payload 回归测试，确保限流、鉴权等流式错误会抛给调用方
 
 ## 下载
 
 | 平台 | 文件 |
 |------|------|
-| Windows | Anything-Analyzer-Setup-3.6.16.exe |
-| macOS (Apple Silicon) | Anything-Analyzer-3.6.16-arm64.dmg |
-| macOS (Intel) | Anything-Analyzer-3.6.16-x64.dmg |
-| Linux | Anything-Analyzer-3.6.16.AppImage |
+| Windows | Anything-Analyzer-Setup-3.6.17.exe |
+| macOS (Apple Silicon) | Anything-Analyzer-3.6.17-arm64.dmg |
+| macOS (Intel) | Anything-Analyzer-3.6.17-x64.dmg |
+| Linux | Anything-Analyzer-3.6.17.AppImage |

package.json

@@ -1,6 +1,6 @@
 {
   "name": "anything-analyzer",
-  "version": "3.6.16",
+  "version": "3.6.17",
   "description": "Universal web protocol analyzer with embedded browser and AI-powered analysis",
   "packageManager": "pnpm@10.24.0",
   "main": "./out/main/index.js",

src/main/ai/llm-router.ts

@@ -635,19 +635,25 @@ export class LLMRouter {
       if (!trimmed || !trimmed.startsWith("data: ")) return;
       const data = trimmed.slice(6);
       if (data === "[DONE]") return;
+      let parsed: any;
       try {
-        const parsed = JSON.parse(data) as any;
-        const chunk = parsed.choices?.[0]?.delta?.content || "";
-        if (chunk) {
-          fullContent += chunk;
-          onChunk(chunk);
-        }
-        if (parsed.usage) {
-          promptTokens = parsed.usage.prompt_tokens;
-          completionTokens = parsed.usage.completion_tokens;
-        }
+        parsed = JSON.parse(data) as any;
       } catch {
-        /* skip */
+        /* skip malformed JSON */
+        return;
+      }
+      if (parsed.error) {
+        const errorMsg = parsed.error.message || "Unknown stream error";
+        throw new Error(`OpenAI stream error: ${errorMsg}`);
+      }
+      const chunk = parsed.choices?.[0]?.delta?.content || "";
+      if (chunk) {
+        fullContent += chunk;
+        onChunk(chunk);
+      }
+      if (parsed.usage) {
+        promptTokens = parsed.usage.prompt_tokens;
+        completionTokens = parsed.usage.completion_tokens;
       }
     };
 

tests/main/ai/llm-router.test.ts

@@ -365,6 +365,22 @@ describe("LLMRouter", () => {
       expect(chunks).toEqual(["final chat chunk"]);
       expect(result.content).toBe("final chat chunk");
     });
+
+    it("should reject when OpenAI chat stream emits an error payload", async () => {
+      fetchSpy.mockResolvedValueOnce(
+        createSSEResponse([
+          {
+            data: '{"error":{"message":"quota exceeded"}}',
+          },
+        ]),
+      );
+
+      const router = new LLMRouter(baseConfig);
+
+      await expect(
+        router.complete([{ role: "user", content: "test" }], () => {}),
+      ).rejects.toThrow("OpenAI stream error: quota exceeded");
+    });
   });
 
   describe("completeAnthropic - streaming", () => {