feat: add image_caption_wait_for_context_order config option

Author: Reisenbug

astrbot/core/config/default.py

@@ -103,6 +103,7 @@
         "fallback_chat_models": [],
         "default_image_caption_provider_id": "",
         "image_caption_prompt": "Please describe the image using Chinese.",
+        "image_caption_wait_for_context_order": True,
         "provider_pool": ["*"],  # "*" 表示使用所有可用的提供者
         "wake_prefix": "",
         "web_search": False,
@@ -2979,6 +2980,11 @@ class ChatProviderTemplate(TypedDict):
                         "_special": "select_provider",
                         "hint": "留空代表不使用，可用于非多模态模型",
                     },
+                    "provider_settings.image_caption_wait_for_context_order": {
+                        "description": "图片转述时等待上下文顺序",
+                        "type": "bool",
+                        "hint": "开启后，同一会话中图片转述完成前，后续消息将等待，以保证上下文顺序正确；关闭后，后续消息立即响应，但上下文中图片描述可能在后续消息之后。",
+                    },
                     "provider_stt_settings.enable": {
                         "description": "启用语音转文本",
                         "type": "bool",

astrbot/core/pipeline/process_stage/method/agent_sub_stages/internal.py

@@ -188,9 +188,12 @@ async def process(
             await call_event_hook(event, EventType.OnWaitingLLMRequestEvent)
 
             if not event.get_extra("provider_request"):
-                await pre_caption_images(
-                    event, self.ctx.plugin_manager.context
-                )
+                plugin_context = self.ctx.plugin_manager.context
+                cfg = plugin_context.get_config(
+                    umo=event.unified_msg_origin
+                ).get("provider_settings", {})
+                if cfg.get("image_caption_wait_for_context_order", True):
+                    await pre_caption_images(event, plugin_context)
 
             async with session_lock_manager.acquire_lock(event.unified_msg_origin):
                 logger.debug("acquired session lock for llm request")