fix: prevent duplicate processing of quoted images by multimodal main providers when no dedicated image caption provider is configured (#8401)

Co-authored-by: C₂₂H₂₅NO₆ <Sisyphbaous-DT-Project@users.noreply.github.com>
Co-authored-by: Soulter <905617992@qq.com>

Author: 3 people

astrbot/core/astr_main_agent.py

@@ -752,6 +752,7 @@ async def _process_quote_message(
     plugin_context: Context,
     quoted_message_settings: QuotedMessageParserSettings = DEFAULT_QUOTED_MESSAGE_SETTINGS,
     config: MainAgentBuildConfig | None = None,
+    main_provider_supports_image: bool = False,
 ) -> None:
     quote = None
     for comp in event.message_obj.message:
@@ -781,13 +782,21 @@ async def _process_quote_message(
                 image_seg = comp
                 break
 
-    if image_seg:
+    if image_seg and main_provider_supports_image:
+        logger.debug(
+            "Skipping quote image captioning because the main provider supports image input."
+        )
+    elif image_seg and not img_cap_prov_id:
+        logger.debug(
+            "No dedicated image caption provider configured. "
+            "Skipping quote image captioning."
+        )
+    elif image_seg:
         try:
             prov = None
             path = None
             compress_path = None
-            if img_cap_prov_id:
-                prov = plugin_context.get_provider_by_id(img_cap_prov_id)
+            prov = plugin_context.get_provider_by_id(img_cap_prov_id)
             if prov is None:
                 prov = plugin_context.get_using_provider(event.unified_msg_origin)
 
@@ -876,18 +885,23 @@ async def _decorate_llm_request(
     req: ProviderRequest,
     plugin_context: Context,
     config: MainAgentBuildConfig,
+    provider: Provider | None = None,
 ) -> None:
     cfg = config.provider_settings or plugin_context.get_config(
         umo=event.unified_msg_origin
     ).get("provider_settings", {})
 
     _apply_prompt_prefix(req, cfg)
 
+    main_provider_supports_image = provider is not None and _provider_supports_modality(
+        provider, "image"
+    )
+
     if req.conversation:
         await _ensure_persona_and_skills(req, cfg, plugin_context, event)
 
         img_cap_prov_id: str = cfg.get("default_image_caption_provider_id") or ""
-        if img_cap_prov_id and req.image_urls:
+        if img_cap_prov_id and req.image_urls and not main_provider_supports_image:
             await _ensure_img_caption(
                 event,
                 req,
@@ -905,6 +919,7 @@ async def _decorate_llm_request(
         plugin_context,
         quoted_message_settings,
         config,
+        main_provider_supports_image=main_provider_supports_image,
     )
 
     tz = config.timezone
@@ -1418,7 +1433,7 @@ async def build_main_agent(
         else:
             return None
 
-    await _decorate_llm_request(event, req, plugin_context, config)
+    await _decorate_llm_request(event, req, plugin_context, config, provider=provider)
 
     await _apply_kb(event, req, plugin_context, config)
 

tests/unit/test_astr_main_agent.py

@@ -1085,6 +1085,61 @@ async def test_build_main_agent_with_images(
 
         assert result is not None
 
+    @pytest.mark.asyncio
+    async def test_build_main_agent_skips_caption_when_main_provider_supports_images(
+        self, mock_event, mock_context, mock_provider
+    ):
+        """Test image-capable chat providers receive quoted images directly."""
+        module = ama
+        mock_image = Image(file="file:///tmp/quoted.jpg")
+        mock_reply = Reply(
+            id="reply-1",
+            chain=[Plain(text="quoted text"), mock_image],
+            sender_nickname="",
+            message_str="quoted text",
+        )
+        mock_event.message_obj.message = [Plain(text="Hello"), mock_reply]
+
+        mock_context.get_provider_by_id.return_value = None
+        mock_context.get_using_provider.return_value = mock_provider
+        mock_context.get_config.return_value = {}
+
+        conv_mgr = mock_context.conversation_manager
+        _setup_conversation_for_build(conv_mgr)
+
+        with (
+            patch("astrbot.core.astr_main_agent.AgentRunner") as mock_runner_cls,
+            patch("astrbot.core.astr_main_agent.AstrAgentContext"),
+            patch.object(
+                Image,
+                "convert_to_file_path",
+                AsyncMock(return_value="/tmp/quoted.jpg"),
+            ),
+        ):
+            mock_runner = MagicMock()
+            mock_runner.reset = AsyncMock()
+            mock_runner_cls.return_value = mock_runner
+
+            result = await module.build_main_agent(
+                event=mock_event,
+                plugin_context=mock_context,
+                config=module.MainAgentBuildConfig(
+                    tool_call_timeout=60,
+                    provider_settings={
+                        "default_image_caption_provider_id": "caption-provider",
+                    },
+                ),
+                provider=mock_provider,
+            )
+
+        assert result is not None
+        assert result.provider_request.image_urls == ["/tmp/quoted.jpg"]
+        assert not any(
+            "Image Caption" in part.text or "<image_caption>" in part.text
+            for part in result.provider_request.extra_user_content_parts
+        )
+        mock_provider.text_chat.assert_not_called()
+
     @pytest.mark.asyncio
     async def test_build_main_agent_uses_image_fallback_provider(
         self, mock_event, mock_context