add retry for xiaoniu api requests

plugins/ai_auto_response/config.toml

@@ -18,6 +18,8 @@ timeout_seconds = 45
 temperature = 0.35
 max_tokens = 120
 stream = true
+max_retries = 3
+retry_delay_seconds = 1.0
 
 [mode]
 group_default_mode = "social"

plugins/ai_auto_response/llm_client.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import json
+import time
 from typing import Dict, List, Optional
 
 import requests
@@ -18,6 +19,8 @@ class LLMClient:
         self.temperature = float(self.config.get("temperature", 0.7))
         self.max_tokens = int(self.config.get("max_tokens", 500))
         self.stream = bool(self.config.get("stream", True))
+        self.max_retries = max(int(self.config.get("max_retries", 3) or 3), 1)
+        self.retry_delay_seconds = float(self.config.get("retry_delay_seconds", 1.0) or 1.0)
         self.last_error = ""
 
     def chat(
@@ -61,25 +64,34 @@ class LLMClient:
         if self.api_key:
             headers["Authorization"] = f"Bearer {self.api_key}"
 
-        try:
+        for attempt in range(1, self.max_retries + 1):
-            if self.stream:
+            try:
-                return self._chat_streaming(payload, headers)
+                if self.stream:
-            response = requests.post(
+                    text = self._chat_streaming(payload, headers)
-                f"{self.base_url}/{self.endpoint}",
+                else:
-                json=payload,
+                    text = self._chat_non_streaming(payload, headers)
-                headers=headers,
+                if text:
-                timeout=self.timeout_seconds,
+                    return text
-            )
+            except Exception as exc:
-            response.raise_for_status()
+                self.last_error = f"request_failed:attempt_{attempt}:{exc}"
-            data = response.json()
+            if attempt < self.max_retries:
-            text = self._extract_text(data)
+                time.sleep(self.retry_delay_seconds * attempt)
-            if text:
+        return ""
-                return text
+
-            self.last_error = f"empty_model_output:{self.model}"
+    def _chat_non_streaming(self, payload: Dict, headers: Dict[str, str]) -> str:
-            return ""
+        response = requests.post(
-        except Exception as exc:
+            f"{self.base_url}/{self.endpoint}",
-            self.last_error = f"request_failed:{exc}"
+            json=payload,
-            return ""
+            headers=headers,
+            timeout=self.timeout_seconds,
+        )
+        response.raise_for_status()
+        data = response.json()
+        text = self._extract_text(data)
+        if text:
+            return text
+        self.last_error = f"empty_model_output:{self.model}"
+        return ""
 
     def _chat_streaming(self, payload: Dict, headers: Dict[str, str]) -> str:
         chunks: List[str] = []