fix：移除冗余代码，丰富replyer log记录，表达方式改为replyer模型

src/config/official_configs.py

@@ -1582,30 +1582,6 @@ class DatabaseConfig(ConfigBase):
     """
 
 
-class MaiSakaConfig(ConfigBase):
-    """MaiSaka 对话系统配置类"""
-
-    __ui_label__ = "MaiSaka"
-    __ui_icon__ = "message-circle"
-    cli_user_name: str = Field(
-        default="用户",
-        json_schema_extra={
-            "x-widget": "input",
-            "x-icon": "user",
-        },
-    )
-    """MaiSaka 使用的用户名称"""
-
-    show_image_path: bool = Field(
-        default=True,
-        json_schema_extra={
-            "x-widget": "switch",
-            "x-icon": "image",
-        },
-    )
-    """是否显示图片本地路径"""
-
-
 class MCPAuthorizationConfig(ConfigBase):
     """MCP HTTP 认证配置。"""
 

src/learners/expression_auto_check_task.py

@@ -8,107 +8,24 @@
 4. 未通过评估的：rejected=1, checked=1
 """
 
+from typing import List
+
 import asyncio
 import random
-from typing import List
 
 from sqlmodel import select
 
-from src.common.data_models.llm_service_data_models import LLMGenerationOptions
 from src.common.database.database import get_db_session
 from src.common.database.database_model import Expression
 from src.common.logger import get_logger
 from src.config.config import global_config
 from src.learners.expression_review_store import get_review_state, set_review_state
-from src.learners.expression_utils import parse_evaluation_response
+from src.learners.expression_utils import check_expression_suitability
 from src.manager.async_task_manager import AsyncTask
-from src.services.llm_service import LLMServiceClient
 
 logger = get_logger("expressor")
 
 
-def create_evaluation_prompt(situation: str, style: str) -> str:
-    """
-    创建评估提示词。
-
-    Args:
-        situation: 情景
-        style: 风格
-
-    Returns:
-        评估提示词
-    """
-    base_criteria = [
-        "表达方式或言语风格是否与使用条件或使用情景匹配",
-        "允许部分语法错误或口语化或缺省出现",
-        "表达方式不能太过特指，需要具有泛用性",
-        "一般不涉及具体的人名或名称",
-    ]
-
-    custom_criteria = global_config.expression.expression_auto_check_custom_criteria
-
-    all_criteria = base_criteria.copy()
-    if custom_criteria:
-        all_criteria.extend(custom_criteria)
-
-    criteria_list = "\n".join([f"{i + 1}. {criterion}" for i, criterion in enumerate(all_criteria)])
-
-    prompt = f"""请评估以下表达方式或语言风格以及使用条件或使用情景是否合适：
-使用条件或使用情景：{situation}
-表达方式或言语风格：{style}
-
-请从以下方面进行评估：
-{criteria_list}
-
-请以 JSON 格式输出评估结果：
-{{
-    "suitable": true/false,
-    "reason": "评估理由（如果不合适，请说明原因）"
-}}
-如果合适，suitable 设为 true；如果不合适，suitable 设为 false，并在 reason 中说明原因。
-请严格按照 JSON 格式输出，不要包含其他内容。"""
-
-    return prompt
-
-
-judge_llm = LLMServiceClient(task_name="utils", request_type="expression_check")
-
-
-async def single_expression_check(situation: str, style: str) -> tuple[bool, str, str | None]:
-    """
-    执行单次 LLM 评估。
-
-    Args:
-        situation: 情景
-        style: 风格
-
-    Returns:
-        (suitable, reason, error) 元组，如果出错则 suitable 为 False，error 包含错误信息
-    """
-    try:
-        prompt = create_evaluation_prompt(situation, style)
-        logger.debug(f"正在评估表达方式: situation={situation}, style={style}")
-
-        generation_result = await judge_llm.generate_response(
-            prompt=prompt,
-            options=LLMGenerationOptions(temperature=0.6, max_tokens=1024),
-        )
-        response = generation_result.response
-        logger.debug(f"LLM响应: {response}")
-
-        evaluation = parse_evaluation_response(response)
-
-        suitable = bool(evaluation.get("suitable", False))
-        reason = str(evaluation.get("reason", "未提供理由"))
-
-        logger.debug(f"评估结果: {'通过' if suitable else '不通过'}")
-        return suitable, reason, None
-
-    except Exception as e:
-        logger.error(f"评估表达方式 (situation={situation}, style={style}) 时出错: {e}")
-        return False, f"评估过程出错: {str(e)}", str(e)
-
-
 class ExpressionAutoCheckTask(AsyncTask):
     """表达方式自动检查定时任务。"""
 
@@ -164,7 +81,7 @@ class ExpressionAutoCheckTask(AsyncTask):
         Returns:
             True 表示通过，False 表示不通过
         """
-        suitable, reason, error = await single_expression_check(
+        suitable, reason, error = await check_expression_suitability(
             expression.situation,
             expression.style,
         )

src/learners/expression_learner.py

@@ -31,11 +31,9 @@ if TYPE_CHECKING:
 logger = get_logger("expressor")
 
 express_learn_model = LLMServiceClient(
-    task_name="utils", request_type="expression.learner"
+    task_name="replyer", request_type="expression.learner"
 )
-summary_model = LLMServiceClient(task_name="utils", request_type="expression.summary")
+summary_model = LLMServiceClient(task_name="replyer", request_type="expression.summary")
-check_model = LLMServiceClient(task_name="utils", request_type="expression.check")
-
 
 def register_expression_hook_specs(registry: HookSpecRegistry) -> List[HookSpec]:
     """注册表达方式系统内置 Hook 规格。

src/learners/expression_utils.py

@@ -12,7 +12,7 @@ from src.services.llm_service import LLMServiceClient
 
 logger = get_logger("expression_utils")
 
-judge_llm = LLMServiceClient(task_name="utils", request_type="expression_check")
+judge_llm = LLMServiceClient(task_name="replyer", request_type="expression_check")
 
 
 def _normalize_repair_json_result(repaired_result: Any) -> str:

src/learners/jargon_explainer_old.py

@@ -1,348 +0,0 @@
-import re
-import time
-from typing import List, Dict, Optional, Any
-
-from src.common.logger import get_logger
-from src.common.database.database_model import Jargon
-from src.common.data_models.llm_service_data_models import LLMGenerationOptions
-from src.services.llm_service import LLMServiceClient
-from src.config.config import global_config
-from src.prompt.prompt_manager import prompt_manager
-from src.learners.jargon_explainer import search_jargon
-from src.learners.learner_utils_old import (
-    is_bot_message,
-    contains_bot_self_name,
-    parse_chat_id_list,
-    chat_id_list_contains,
-)
-
-logger = get_logger("jargon")
-
-
-class JargonExplainer:
-    """黑话解释器，用于在回复前识别和解释上下文中的黑话"""
-
-    def __init__(self, chat_id: str) -> None:
-        self.chat_id = chat_id
-        self.llm = LLMServiceClient(
-            task_name="utils",
-            request_type="jargon.explain",
-        )
-
-    def match_jargon_from_messages(self, messages: List[Any]) -> List[Dict[str, str]]:
-        """
-        通过直接匹配数据库中的jargon字符串来提取黑话
-
-        Args:
-            messages: 消息列表
-
-        Returns:
-            List[Dict[str, str]]: 提取到的黑话列表，每个元素包含content
-        """
-        start_time = time.time()
-
-        if not messages:
-            return []
-
-        # 收集所有消息的文本内容
-        message_texts: List[str] = []
-        for msg in messages:
-            # 跳过机器人自己的消息
-            if is_bot_message(msg):
-                continue
-
-            msg_text = (
-                getattr(msg, "display_message", None) or getattr(msg, "processed_plain_text", None) or ""
-            ).strip()
-            if msg_text:
-                message_texts.append(msg_text)
-
-        if not message_texts:
-            return []
-
-        # 合并所有消息文本
-        combined_text = " ".join(message_texts)
-
-        # 查询所有有meaning的jargon记录
-        query = Jargon.select().where((Jargon.meaning.is_null(False)) & (Jargon.meaning != ""))
-
-        # 根据all_global配置决定查询逻辑
-        if global_config.expression.all_global_jargon:
-            # 开启all_global：只查询is_global=True的记录
-            query = query.where(Jargon.is_global)
-        else:
-            # 关闭all_global：查询is_global=True或chat_id列表包含当前chat_id的记录
-            # 这里先查询所有，然后在Python层面过滤
-            pass
-
-        # 按count降序排序，优先匹配出现频率高的
-        query = query.order_by(Jargon.count.desc())
-
-        # 执行查询并匹配
-        matched_jargon: Dict[str, Dict[str, str]] = {}
-        query_time = time.time()
-
-        for jargon in query:
-            content = jargon.content or ""
-            if not content or not content.strip():
-                continue
-
-            # 跳过包含机器人昵称的词条
-            if contains_bot_self_name(content):
-                continue
-
-            # 检查chat_id（如果all_global=False）
-            if not global_config.expression.all_global_jargon:
-                if jargon.is_global:
-                    # 全局黑话，包含
-                    pass
-                else:
-                    # 检查chat_id列表是否包含当前chat_id
-                    chat_id_list = parse_chat_id_list(jargon.chat_id)
-                    if not chat_id_list_contains(chat_id_list, self.chat_id):
-                        continue
-
-            # 在文本中查找匹配（大小写不敏感）
-            pattern = re.escape(content)
-            # 使用单词边界或中文字符边界来匹配，避免部分匹配
-            # 对于中文，使用Unicode字符类；对于英文，使用单词边界
-            if re.search(r"[\u4e00-\u9fff]", content):
-                # 包含中文，使用更宽松的匹配
-                search_pattern = pattern
-            else:
-                # 纯英文/数字，使用单词边界
-                search_pattern = r"\b" + pattern + r"\b"
-
-            if re.search(search_pattern, combined_text, re.IGNORECASE):
-                # 找到匹配，记录（去重）
-                if content not in matched_jargon:
-                    matched_jargon[content] = {"content": content}
-
-        match_time = time.time()
-        total_time = match_time - start_time
-        query_duration = query_time - start_time
-        match_duration = match_time - query_time
-
-        logger.debug(
-            f"黑话匹配完成: 查询耗时 {query_duration:.3f}s, 匹配耗时 {match_duration:.3f}s, "
-            f"总耗时 {total_time:.3f}s, 匹配到 {len(matched_jargon)} 个黑话"
-        )
-
-        return list(matched_jargon.values())
-
-    async def explain_jargon(self, messages: List[Any], chat_context: str) -> Optional[str]:
-        """
-        解释上下文中的黑话
-
-        Args:
-            messages: 消息列表
-            chat_context: 聊天上下文的文本表示
-
-        Returns:
-            Optional[str]: 黑话解释的概括文本，如果没有黑话则返回None
-        """
-        if not messages:
-            return None
-
-        # 直接匹配方式：从数据库中查询jargon并在消息中匹配
-        jargon_entries = self.match_jargon_from_messages(messages)
-
-        if not jargon_entries:
-            return None
-
-        # 去重（按content）
-        unique_jargon: Dict[str, Dict[str, str]] = {}
-        for entry in jargon_entries:
-            content = entry["content"]
-            if content not in unique_jargon:
-                unique_jargon[content] = entry
-
-        jargon_list = list(unique_jargon.values())
-        logger.info(f"从上下文中提取到 {len(jargon_list)} 个黑话: {[j['content'] for j in jargon_list]}")
-
-        # 查询每个黑话的含义
-        jargon_explanations: List[str] = []
-        for entry in jargon_list:
-            content = entry["content"]
-
-            # 根据是否开启全局黑话，决定查询方式
-            if global_config.expression.all_global_jargon:
-                # 开启全局黑话：查询所有is_global=True的记录
-                results = search_jargon(
-                    keyword=content,
-                    chat_id=None,  # 不指定chat_id，查询全局黑话
-                    limit=1,
-                    case_sensitive=False,
-                    fuzzy=False,  # 精确匹配
-                )
-            else:
-                # 关闭全局黑话：优先查询当前聊天或全局的黑话
-                results = search_jargon(
-                    keyword=content,
-                    chat_id=self.chat_id,
-                    limit=1,
-                    case_sensitive=False,
-                    fuzzy=False,  # 精确匹配
-                )
-
-            if results and len(results) > 0:
-                meaning = results[0].get("meaning", "").strip()
-                if meaning:
-                    jargon_explanations.append(f"- {content}: {meaning}")
-                else:
-                    logger.info(f"黑话 {content} 没有找到含义")
-            else:
-                logger.info(f"黑话 {content} 未在数据库中找到")
-
-        if not jargon_explanations:
-            logger.info("没有找到任何黑话的含义，跳过解释")
-            return None
-
-        # 拼接所有黑话解释
-        explanations_text = "\n".join(jargon_explanations)
-
-        # 使用LLM概括黑话解释
-        prompt_of_summarize = prompt_manager.get_prompt("jargon_explainer_summarize")
-        prompt_of_summarize.add_context("chat_context", lambda _: chat_context)
-        prompt_of_summarize.add_context("jargon_explanations", lambda _: explanations_text)
-        summarize_prompt = await prompt_manager.render_prompt(prompt_of_summarize)
-
-        summary_result = await self.llm.generate_response(
-            summarize_prompt, options=LLMGenerationOptions(temperature=0.3)
-        )
-        summary = summary_result.response
-        if not summary:
-            # 如果LLM概括失败，直接返回原始解释
-            return f"上下文中的黑话解释：\n{explanations_text}"
-
-        summary = summary.strip()
-        if not summary:
-            return f"上下文中的黑话解释：\n{explanations_text}"
-
-        return summary
-
-
-async def explain_jargon_in_context(chat_id: str, messages: List[Any], chat_context: str) -> Optional[str]:
-    """
-    解释上下文中的黑话（便捷函数）
-
-    Args:
-        chat_id: 聊天ID
-        messages: 消息列表
-        chat_context: 聊天上下文的文本表示
-
-    Returns:
-        Optional[str]: 黑话解释的概括文本，如果没有黑话则返回None
-    """
-    explainer = JargonExplainer(chat_id)
-    return await explainer.explain_jargon(messages, chat_context)
-
-
-def match_jargon_from_text(chat_text: str, chat_id: str) -> List[str]:
-    """直接在聊天文本中匹配已知的jargon，返回出现过的黑话列表
-
-    Args:
-        chat_text: 要匹配的聊天文本
-        chat_id: 聊天ID
-
-    Returns:
-        List[str]: 匹配到的黑话列表
-    """
-    if not chat_text or not chat_text.strip():
-        return []
-
-    query = Jargon.select().where((Jargon.meaning.is_null(False)) & (Jargon.meaning != ""))
-    if global_config.expression.all_global_jargon:
-        query = query.where(Jargon.is_global)
-
-    query = query.order_by(Jargon.count.desc())
-
-    matched: Dict[str, None] = {}
-
-    for jargon in query:
-        content = (jargon.content or "").strip()
-        if not content:
-            continue
-
-        if not global_config.expression.all_global_jargon and not jargon.is_global:
-            chat_id_list = parse_chat_id_list(jargon.chat_id)
-            if not chat_id_list_contains(chat_id_list, chat_id):
-                continue
-
-        pattern = re.escape(content)
-        if re.search(r"[\u4e00-\u9fff]", content):
-            search_pattern = pattern
-        else:
-            search_pattern = r"\b" + pattern + r"\b"
-
-        if re.search(search_pattern, chat_text, re.IGNORECASE):
-            matched[content] = None
-
-    logger.info(f"匹配到 {len(matched)} 个黑话")
-
-    return list(matched.keys())
-
-
-async def retrieve_concepts_with_jargon(concepts: List[str], chat_id: str) -> str:
-    """对概念列表进行jargon检索
-
-    Args:
-        concepts: 概念列表
-        chat_id: 聊天ID
-
-    Returns:
-        str: 检索结果字符串
-    """
-    if not concepts:
-        return ""
-
-    results = []
-    exact_matches = []  # 收集所有精确匹配的概念
-    for concept in concepts:
-        concept = concept.strip()
-        if not concept:
-            continue
-
-        # 先尝试精确匹配
-        jargon_results = search_jargon(keyword=concept, chat_id=chat_id, limit=10, case_sensitive=False, fuzzy=False)
-
-        is_fuzzy_match = False
-
-        # 如果精确匹配未找到，尝试模糊搜索
-        if not jargon_results:
-            jargon_results = search_jargon(keyword=concept, chat_id=chat_id, limit=10, case_sensitive=False, fuzzy=True)
-            is_fuzzy_match = True
-
-        if jargon_results:
-            # 找到结果
-            if is_fuzzy_match:
-                # 模糊匹配
-                output_parts = [f"未精确匹配到'{concept}'"]
-                for result in jargon_results:
-                    found_content = result.get("content", "").strip()
-                    meaning = result.get("meaning", "").strip()
-                    if found_content and meaning:
-                        output_parts.append(f"找到 '{found_content}' 的含义为：{meaning}")
-                results.append("\n".join(output_parts))  # 换行分隔每个jargon解释
-                logger.info(f"在jargon库中找到匹配（模糊搜索）: {concept}，找到{len(jargon_results)}条结果")
-            else:
-                # 精确匹配
-                output_parts = []
-                for result in jargon_results:
-                    meaning = result.get("meaning", "").strip()
-                    if meaning:
-                        output_parts.append(f"'{concept}' 为黑话或者网络简写，含义为：{meaning}")
-                # 换行分隔每个jargon解释
-                results.append("\n".join(output_parts) if len(output_parts) > 1 else output_parts[0])
-                exact_matches.append(concept)  # 收集精确匹配的概念，稍后统一打印
-        else:
-            # 未找到，不返回占位信息，只记录日志
-            logger.info(f"在jargon库中未找到匹配: {concept}")
-
-    # 合并所有精确匹配的日志
-    if exact_matches:
-        logger.info(f"找到黑话: {', '.join(exact_matches)}，共找到{len(exact_matches)}条结果")
-
-    if results:
-        return "你了解以下词语可能的含义：\n" + "\n".join(results) + "\n"
-    return ""

src/learners/jargon_miner.py

@@ -23,8 +23,8 @@ from .expression_utils import is_single_char_jargon
 
 logger = get_logger("jargon")
 
-llm_extract = LLMServiceClient(task_name="utils", request_type="jargon.extract")
+llm_extract = LLMServiceClient(task_name="replyer", request_type="jargon.extract")
-llm_inference = LLMServiceClient(task_name="utils", request_type="jargon.inference")
+llm_inference = LLMServiceClient(task_name="replyer", request_type="jargon.inference")
 
 
 class JargonEntry(TypedDict):

src/maisaka/builtin_tool/reply.py

@@ -1,6 +1,6 @@
 """reply 内置工具。"""
 
-from typing import Optional
+from typing import Any, Optional
 
 import traceback
 
@@ -75,6 +75,25 @@ def _build_monitor_metadata(reply_result: ReplyGenerationResult) -> dict[str, ob
     return {}
 
 
+def _build_send_result(
+    *,
+    index: int,
+    segment: str,
+    set_quote: bool,
+    success: bool,
+    message_id: str = "",
+) -> dict[str, Any]:
+    """构建分段回复的轻量发送结果。"""
+
+    return {
+        "index": index,
+        "segment": segment,
+        "set_quote": set_quote,
+        "success": success,
+        "message_id": message_id,
+    }
+
+
 async def handle_tool(
     tool_ctx: BuiltinToolRuntimeContext,
     invocation: ToolInvocation,
@@ -165,19 +184,29 @@ async def handle_tool(
     reply_segments = tool_ctx.post_process_reply_text(reply_text)
     combined_reply_text = "".join(reply_segments)
     sent_message_ids: list[str] = []
+    send_results: list[dict[str, Any]] = []
     try:
         sent = False
         if tool_ctx.runtime.chat_stream.platform == CLI_PLATFORM_NAME:
-            for segment in reply_segments:
+            for index, segment in enumerate(reply_segments):
                 render_cli_message(segment)
+                send_results.append(
+                    _build_send_result(
+                        index=index,
+                        segment=segment,
+                        set_quote=effective_set_quote if index == 0 else False,
+                        success=True,
+                    )
+                )
             sent = True
         else:
             for index, segment in enumerate(reply_segments):
+                segment_set_quote = effective_set_quote if index == 0 else False
                 sent_message = await send_service.text_to_stream_with_message(
                     text=segment,
                     stream_id=tool_ctx.runtime.session_id,
-                    set_reply=effective_set_quote if index == 0 else False,
+                    set_reply=segment_set_quote,
-                    reply_message=target_message if effective_set_quote and index == 0 else None,
+                    reply_message=target_message if segment_set_quote else None,
                     selected_expressions=reply_result.selected_expression_ids or None,
                     typing=index > 0,
                     sync_to_maisaka_history=True,
@@ -185,10 +214,27 @@ async def handle_tool(
                 )
                 sent = sent_message is not None
                 if not sent:
+                    send_results.append(
+                        _build_send_result(
+                            index=index,
+                            segment=segment,
+                            set_quote=segment_set_quote,
+                            success=False,
+                        )
+                    )
                     break
                 sent_message_id = str(getattr(sent_message, "message_id", "") or "").strip()
                 if sent_message_id:
                     sent_message_ids.append(sent_message_id)
+                send_results.append(
+                    _build_send_result(
+                        index=index,
+                        segment=segment,
+                        set_quote=segment_set_quote,
+                        success=True,
+                        message_id=sent_message_id,
+                    )
+                )
     except Exception:
         logger.exception(
             f"{tool_ctx.runtime.log_prefix} 发送文字消息时发生异常，目标消息编号={target_message_id}"
@@ -208,6 +254,7 @@ async def handle_tool(
                 "set_quote": set_quote,
                 "effective_set_quote": effective_set_quote,
                 "reply_segments": reply_segments,
+                "send_results": send_results,
             },
             metadata=reply_metadata,
         )
@@ -219,7 +266,10 @@ async def handle_tool(
         tool_ctx.append_guided_reply_to_chat_history(combined_reply_text)
     tool_ctx.runtime._record_reply_sent()
     reply_metadata["sent_message_ids"] = sent_message_ids
-    await tool_ctx.runtime.track_reply_effect(
+    reply_metadata["send_results"] = send_results
+    track_reply_effect = getattr(tool_ctx.runtime, "track_reply_effect", None)
+    if track_reply_effect is not None:
+        await track_reply_effect(
             tool_call_id=invocation.call_id,
             target_message=target_message,
             set_quote=effective_set_quote,
@@ -227,6 +277,13 @@ async def handle_tool(
             reply_segments=reply_segments,
             planner_reasoning=latest_thought,
             reference_info=reference_info,
+            tool_context={
+                "tool_name": invocation.tool_name,
+                "call_id": invocation.call_id,
+                "arguments": dict(invocation.arguments or {}),
+                "reasoning": latest_thought,
+            },
+            send_results=send_results,
             reply_metadata=reply_metadata,
             replyer_context_messages=replyer_chat_history,
         )
@@ -239,6 +296,7 @@ async def handle_tool(
             "effective_set_quote": effective_set_quote,
             "reply_text": combined_reply_text,
             "reply_segments": reply_segments,
+            "send_results": send_results,
             "target_user_name": target_user_name,
         },
         metadata=reply_metadata,

src/maisaka/reply_effect/models.py

@@ -50,6 +50,8 @@ class ReplySnapshot:
     reply_segments: List[str]
     planner_reasoning: str
     reference_info: str
+    tool_context: Dict[str, Any] = field(default_factory=dict)
+    send_results: List[Dict[str, Any]] = field(default_factory=list)
     reply_metadata: Dict[str, Any] = field(default_factory=dict)
 
 

src/maisaka/reply_effect/tracker.py

@@ -67,6 +67,8 @@ class ReplyEffectTracker:
         reply_segments: List[str],
         planner_reasoning: str,
         reference_info: str,
+        tool_context: Dict[str, Any] | None = None,
+        send_results: List[Dict[str, Any]] | None = None,
         reply_metadata: Dict[str, Any] | None = None,
         context_snapshot: List[Dict[str, Any]] | None = None,
     ) -> ReplyEffectRecord:
@@ -88,6 +90,8 @@ class ReplyEffectTracker:
                 reply_segments=list(reply_segments),
                 planner_reasoning=planner_reasoning,
                 reference_info=reference_info,
+                tool_context=dict(tool_context or {}),
+                send_results=list(send_results or []),
                 reply_metadata=dict(reply_metadata or {}),
             ),
             target_user=UserSnapshot(

src/maisaka/runtime.py

@@ -296,6 +296,8 @@ class MaisakaHeartFlowChatting:
         reply_segments: list[str],
         planner_reasoning: str,
         reference_info: str,
+        tool_context: Optional[dict[str, Any]] = None,
+        send_results: Optional[list[dict[str, Any]]] = None,
         reply_metadata: Optional[dict[str, Any]] = None,
         replyer_context_messages: Optional[Sequence[LLMContextMessage]] = None,
     ) -> None:
@@ -322,6 +324,8 @@ class MaisakaHeartFlowChatting:
                 reply_segments=reply_segments,
                 planner_reasoning=planner_reasoning,
                 reference_info=reference_info,
+                tool_context=tool_context,
+                send_results=send_results,
                 reply_metadata=enriched_reply_metadata,
                 context_snapshot=context_snapshot,
             )