完善成员画像插件的日/周/月分层提取与已结束日期处理逻辑

- 将成员画像能力进一步收敛到插件内部，强化按群启用、后台异步刷新、后台查看的完整链路
- 新增群维度按日批量提取能力：以群为单位按天处理一次，统一提取当天活跃成员的日级画像摘要
- 日级画像输出扩展为更适合长期累计的结构化信号，补充身份线索、技能信号、家庭线索、阶段线索、价值偏好、群内角色、决策风格等字段
- 优化提示词设计，明确要求优先提取可复用、可累计、可验证的行为线索，减少一次性情绪和短期噪声对长期画像的干扰
- 打通日 -> 周 -> 月 -> 最终画像 的分层汇总链路，让后续月度画像直接消费日/周级结构化摘要，而不是重复回扫长窗口原始消息
- 新增/完善画像融合策略：identity_traits、skill_profile、family_profile、life_stage_profile、value_profile 也纳入长期分数累计，不再仅依赖最近一次结果覆盖旧结果
- 将活跃群、活跃成员、辅助消息样本等口径统一调整为只处理已结束日期，避免当天未完结数据进入画像计算
- 调整日级批处理逻辑，默认只处理昨天及更早日期，确保不会处理当天消息
- 修复重复执行时仍然先调用 AI 再跳过的问题，改为先检查当天候选成员是否已完成生成，全部已存在时直接跳过，减少无效 AI 请求和耗时
- 增加群日批处理、周摘要、月摘要、群刷新进度等日志，方便后台定位当前刷新到哪些群、哪些成员、进度如何
- 丰富后台画像展示字段，支持查看更完整的长期画像维度与摘要统计
- 更新插件配置默认值，收敛为近 60 天启动窗口、每日滚动处理与群级日摘要模式
- 补充 message_storage 读取能力，支持按群按日提取消息，为群日批量画像与后续周期汇总提供底层数据支撑

plugins/member_context/config.toml

@@ -12,21 +12,22 @@ request_timeout = 60
 sample_days = 30
 sample_message_limit = 80
 refresh_limit_per_member = 200
-long_term_days = 365
+long_term_days = 60
 long_term_message_limit = 600
-bootstrap_days = 365
+bootstrap_days = 60
 bootstrap_message_limit = 600
 incremental_message_limit = 80
 incremental_recent_days = 7
 recalibration_days = 30
 daily_message_limit = 120
 daily_digest_min_messages = 6
-max_daily_digests_per_run = 45
+max_daily_digests_per_run = 0
 weekly_digest_limit = 16
 monthly_digest_limit = 12
 final_daily_limit = 8
 final_weekly_limit = 6
 final_monthly_limit = 6
+group_digest_days = 1
 ai_min_member_messages = 12
 active_member_hours = 72
 min_member_messages = 3

plugins/member_context/digest_service.py

@@ -12,6 +12,7 @@ from db.contacts_db import ContactsDBOperator
 from db.member_digest_db import MemberDigestDBOperator
 from db.message_storage import MessageStorageDB
 from plugins.member_context.prompt_builder import MemberContextPromptBuilder
+from utils.compress_chat_data import compress_chat_data
 
 
 class MemberDigestService:
@@ -37,28 +38,47 @@ class MemberDigestService:
         self.bootstrap_days = int(profile_config.get("bootstrap_days", 365))
         self.daily_message_limit = int(profile_config.get("daily_message_limit", 120))
         self.daily_digest_min_messages = int(profile_config.get("daily_digest_min_messages", 6))
-        self.max_daily_digests_per_run = int(profile_config.get("max_daily_digests_per_run", 45))
+        self.max_daily_digests_per_run = int(profile_config.get("max_daily_digests_per_run", 0))
         self.weekly_digest_limit = int(profile_config.get("weekly_digest_limit", 16))
         self.monthly_digest_limit = int(profile_config.get("monthly_digest_limit", 12))
         self.final_daily_limit = int(profile_config.get("final_daily_limit", 8))
         self.final_weekly_limit = int(profile_config.get("final_weekly_limit", 6))
         self.final_monthly_limit = int(profile_config.get("final_monthly_limit", 6))
+        self.group_digest_days = int(profile_config.get("group_digest_days", 1))
+
+    def ensure_recent_group_daily_digests(self, chatroom_id: str, days: Optional[int] = None,
+                                          force: bool = False) -> Dict:
+        days = days or self.group_digest_days
+        built_daily = 0
+        touched_members = set()
+
+        for offset in range(days):
+            target_date = (datetime.now() - timedelta(days=offset + 1)).strftime("%Y-%m-%d")
+            digests = self._build_group_daily_digests(chatroom_id, target_date, force=force)
+            for digest in digests:
+                self.digest_db.save_digest(digest)
+                built_daily += 1
+                touched_members.add(digest.get("wxid"))
+
+        return {
+            "built_daily": built_daily,
+            "touched_members": list(touched_members),
+        }
 
     def ensure_member_digest_pipeline(self, chatroom_id: str, wxid: str, force: bool = False) -> Dict:
         member = self.contacts_db.get_chatroom_member_info(chatroom_id, wxid) or {}
         display_name = member.get("display_name") or member.get("nick_name") or wxid
 
-        active_dates = self.message_db.get_member_active_dates(chatroom_id, wxid, days=self.bootstrap_days)
-        if not active_dates:
+        daily_digests = self.digest_db.list_digests(chatroom_id, wxid, "daily", limit=400)
+        if not daily_digests:
             return {
                 "display_name": display_name,
                 "daily_digests": [],
                 "weekly_digests": [],
                 "monthly_digests": [],
-                "stats": {"daily": 0, "weekly": 0, "monthly": 0, "active_days": 0},
+                "stats": {"daily": 0, "weekly": 0, "monthly": 0, "active_days": 0, "built_daily": 0},
             }
 
-        built_daily = self._ensure_daily_digests(chatroom_id, wxid, display_name, active_dates, force=force)
         built_weekly = self._ensure_weekly_digests(chatroom_id, wxid, display_name, force=force)
         built_monthly = self._ensure_monthly_digests(chatroom_id, wxid, display_name, force=force)
 
@@ -75,45 +95,37 @@ class MemberDigestService:
                 "daily": len(daily_digests),
                 "weekly": len(weekly_digests),
                 "monthly": len(monthly_digests),
-                "active_days": len(active_dates),
-                "built_daily": built_daily,
+                "active_days": len(self.digest_db.list_digest_keys(chatroom_id, wxid, "daily")),
+                "built_daily": 0,
                 "built_weekly": built_weekly,
                 "built_monthly": built_monthly,
             },
         }
 
-    def _ensure_daily_digests(self, chatroom_id: str, wxid: str, display_name: str,
-                              active_dates: List[Dict], force: bool = False) -> int:
-        existing_keys = set(self.digest_db.list_digest_keys(chatroom_id, wxid, "daily"))
-        built = 0
-        processed = 0
-        sorted_dates = sorted(active_dates, key=lambda item: str(item.get("message_date")))
-        current_day = datetime.now().strftime("%Y-%m-%d")
-
-        for item in sorted_dates:
-            period_key = str(item.get("message_date"))
-            msg_count = int(item.get("msg_count", 0))
-            if msg_count < self.daily_digest_min_messages:
-                continue
-            if not force and period_key in existing_keys and period_key != current_day:
-                continue
-            messages = self.message_db.get_member_messages_on_date(
-                chatroom_id, wxid, period_key, limit=self.daily_message_limit
-            )
-            if len(messages) < self.daily_digest_min_messages:
-                continue
-            digest = self._build_daily_digest(chatroom_id, wxid, display_name, period_key, messages)
-            if digest:
-                self.digest_db.save_digest(digest)
-                built += 1
-                processed += 1
-                self.LOG.info(
-                    f"[成员交互摘要][日摘要] 完成: group={chatroom_id}, wxid={wxid}, "
-                    f"date={period_key}, messages={len(messages)}"
-                )
-            if not force and processed >= self.max_daily_digests_per_run:
-                break
-        return built
+    @staticmethod
+    def _normalize_profile_item(item: Dict) -> Dict:
+        normalized = {}
+        list_keys = {
+            "topics", "identity_clues", "skill_signals", "family_signals", "life_stage_signals",
+            "value_preferences", "habit_signals", "engagement_traits", "reply_taboos",
+            "representative_messages", "stable_topics", "identity_traits", "skill_profile",
+            "family_profile", "life_stage_profile", "value_profile", "stable_traits",
+            "habit_patterns", "reply_preferences", "long_term_topics", "long_term_reply_preferences",
+            "phase_state", "recent_state"
+        }
+        for key, value in item.items():
+            if key in list_keys:
+                if isinstance(value, list):
+                    normalized[key] = [str(v).strip() for v in value if str(v).strip()]
+                elif value:
+                    normalized[key] = [str(value).strip()]
+                else:
+                    normalized[key] = []
+            elif isinstance(value, (int, float)):
+                normalized[key] = value
+            else:
+                normalized[key] = str(value).strip()
+        return normalized
 
     def _ensure_weekly_digests(self, chatroom_id: str, wxid: str, display_name: str, force: bool = False) -> int:
         daily_digests = self.digest_db.list_digests(chatroom_id, wxid, "daily", limit=400)
@@ -123,7 +135,7 @@ class MemberDigestService:
             grouped[week_key].append(item)
 
         existing_keys = set(self.digest_db.list_digest_keys(chatroom_id, wxid, "weekly"))
-        current_week_key, _, _ = self._week_period_bounds(datetime.now().strftime("%Y-%m-%d"))
+        current_week_key, _, _ = self._week_period_bounds((datetime.now() - timedelta(days=1)).strftime("%Y-%m-%d"))
         built = 0
         for week_key, items in sorted(grouped.items()):
             if len(items) < 2:
@@ -151,7 +163,7 @@ class MemberDigestService:
             grouped[month_key].append(item)
 
         existing_keys = set(self.digest_db.list_digest_keys(chatroom_id, wxid, "monthly"))
-        current_month_key, _, _ = self._month_period_bounds(datetime.now().strftime("%Y-%m-%d"))
+        current_month_key, _, _ = self._month_period_bounds((datetime.now() - timedelta(days=1)).strftime("%Y-%m-%d"))
         built = 0
         for month_key, items in sorted(grouped.items()):
             if len(items) < 2:
@@ -171,6 +183,83 @@ class MemberDigestService:
                 )
         return built
 
+    def _build_group_daily_digests(self, chatroom_id: str, digest_date: str, force: bool = False) -> List[Dict]:
+        members = self.contacts_db.get_chatroom_member_list(chatroom_id) or []
+        member_name_map = {}
+        for member in members:
+            wxid = member.get("wxid")
+            if not wxid:
+                continue
+            member_name_map[wxid] = member.get("display_name") or member.get("nick_name") or wxid
+
+        messages = self.message_db.get_member_messages_for_group_date(chatroom_id, digest_date)
+        if not messages:
+            return []
+
+        sender_messages = defaultdict(list)
+        for msg in messages:
+            wxid = msg.get("sender")
+            if not wxid:
+                continue
+            sender_messages[wxid].append(msg)
+
+        candidate_wxids = [
+            wxid for wxid, items in sender_messages.items()
+            if len(items) >= self.daily_digest_min_messages
+        ]
+        if not candidate_wxids:
+            return []
+
+        pending_wxids = []
+        for wxid in candidate_wxids:
+            if not force and self.digest_db.get_digest(chatroom_id, wxid, "daily", digest_date):
+                continue
+            pending_wxids.append(wxid)
+        if not pending_wxids:
+            self.LOG.info(
+                f"[成员交互摘要][群日批处理] 跳过: group={chatroom_id}, date={digest_date}, "
+                f"reason=all_candidates_already_built, candidates={len(candidate_wxids)}"
+            )
+            return []
+
+        member_labels = [f"{wxid} | {member_name_map.get(wxid, wxid)}" for wxid in pending_wxids]
+        compact_chat = self._format_group_messages_optimized(messages, member_name_map)
+        try:
+            compact_chat = compress_chat_data(compact_chat)
+        except Exception as e:
+            self.LOG.warning(f"[成员交互摘要] 压缩群日消息失败: group={chatroom_id}, date={digest_date}, error={e}")
+
+        parsed_members = self._request_group_daily_json(chatroom_id, digest_date, member_labels, compact_chat)
+        parsed_map = {item.get("wxid"): item for item in parsed_members if item.get("wxid")} if parsed_members else {}
+
+        digests = []
+        for wxid in pending_wxids:
+            parsed = parsed_map.get(wxid) or self._build_daily_digest_fallback(sender_messages.get(wxid, []))
+            if not parsed:
+                continue
+            parsed = self._normalize_profile_item(parsed)
+            digests.append({
+                "chatroom_id": chatroom_id,
+                "wxid": wxid,
+                "digest_type": "daily",
+                "period_key": digest_date,
+                "period_start": f"{digest_date} 00:00:00",
+                "period_end": f"{digest_date} 23:59:59",
+                "display_name": member_name_map.get(wxid, wxid),
+                "source_count": len(sender_messages.get(wxid, [])),
+                "summary_text": parsed.get("summary_text", ""),
+                "structured": parsed,
+                "meta": {
+                    "source_type": "group_daily_messages",
+                    "representative_messages": parsed.get("representative_messages", []),
+                },
+            })
+        self.LOG.info(
+            f"[成员交互摘要][群日批处理] 完成: group={chatroom_id}, date={digest_date}, "
+            f"candidates={len(candidate_wxids)}, pending={len(pending_wxids)}, built={len(digests)}"
+        )
+        return digests
+
     def _build_daily_digest(self, chatroom_id: str, wxid: str, display_name: str,
                             digest_date: str, messages: List[Dict]) -> Optional[Dict]:
         prompt = MemberContextPromptBuilder.build_daily_digest_prompt(
@@ -256,6 +345,39 @@ class MemberDigestService:
             self.LOG.warning(f"[成员交互摘要][AI] 摘要请求失败: group={chatroom_id}, wxid={wxid}, tag={tag}, error={e}")
             return None
 
+    def _request_group_daily_json(self, chatroom_id: str, digest_date: str,
+                                  member_labels: List[str], compressed_chat: str) -> List[Dict]:
+        if not self.ai_enabled or not self.ai_base_url or not self.ai_api_key:
+            return []
+        prompt = MemberContextPromptBuilder.build_group_daily_digest_prompt(
+            chatroom_id, digest_date, member_labels, compressed_chat
+        )
+        headers = {
+            "Authorization": f"Bearer {self.ai_api_key}",
+            "Content-Type": "application/json",
+        }
+        payload = {
+            "inputs": {"query": prompt},
+            "response_mode": "blocking",
+            "user": f"member-digest:{chatroom_id}:group-daily:{digest_date}",
+        }
+        url = f"{self.ai_base_url}/{self.ai_endpoint}"
+        try:
+            self.LOG.info(
+                f"[成员交互摘要][AI] 发起群日批量摘要请求: group={chatroom_id}, "
+                f"date={digest_date}, members={len(member_labels)}"
+            )
+            response = requests.post(url, headers=headers, json=payload, timeout=self.ai_timeout)
+            response.raise_for_status()
+            data = response.json()
+            parsed = self._parse_group_daily_answer(data.get("answer", ""))
+            return parsed
+        except Exception as e:
+            self.LOG.warning(
+                f"[成员交互摘要][AI] 群日批量摘要失败: group={chatroom_id}, date={digest_date}, error={e}"
+            )
+            return []
+
     def _parse_ai_answer(self, answer: str) -> Optional[Dict]:
         if not answer:
             return None
@@ -277,6 +399,29 @@ class MemberDigestService:
                 normalized[key] = str(value).strip()
         return normalized
 
+    def _parse_group_daily_answer(self, answer: str) -> List[Dict]:
+        if not answer:
+            return []
+        text = answer.strip()
+        match = re.search(r"\{.*\}", text, re.S)
+        if match:
+            text = match.group(0)
+        try:
+            parsed = json.loads(text)
+        except Exception:
+            return []
+        members = parsed.get("members", [])
+        if not isinstance(members, list):
+            return []
+        normalized = []
+        for item in members:
+            if not isinstance(item, dict):
+                continue
+            normalized_item = self._normalize_profile_item(item)
+            if normalized_item.get("wxid"):
+                normalized.append(normalized_item)
+        return normalized
+
     def _build_daily_digest_fallback(self, messages: List[Dict]) -> Optional[Dict]:
         if not messages:
             return None
@@ -288,11 +433,18 @@ class MemberDigestService:
         message_pattern = "短句居多" if avg_len <= 16 else "表达较完整" if avg_len >= 35 else "表达中等长度"
         return {
             "topics": [],
+            "identity_clues": [],
+            "skill_signals": [],
+            "family_signals": [],
+            "life_stage_signals": [],
+            "value_preferences": [],
             "interaction_style": "自然跟随式互动",
             "message_pattern": message_pattern,
             "response_style_hint": "保持简洁自然，先回应核心点",
             "habit_signals": [],
             "engagement_traits": [],
+            "decision_style": "",
+            "social_role": "",
             "reply_taboos": [],
             "temperament_signal": "当天样本有限，暂以中性沟通观察为主",
             "summary_text": f"当日消息约{len(messages)}条，{message_pattern}。",
@@ -332,9 +484,16 @@ class MemberDigestService:
         if digest_type == "weekly":
             return {
                 "stable_topics": top_topics,
+                "identity_traits": [],
+                "skill_profile": [],
+                "family_profile": [],
+                "life_stage_profile": [],
+                "value_profile": [],
                 "stable_traits": top_traits,
                 "habit_patterns": top_habits,
                 "reply_preferences": top_reply,
+                "group_role": "",
+                "decision_profile": "",
                 "recent_state": top_topics[:3],
                 "temperament_tendency": temperament,
                 "summary_text": "本周沟通特征已按重复信号汇总。",
@@ -343,15 +502,53 @@ class MemberDigestService:
 
         return {
             "long_term_topics": top_topics,
+            "identity_traits": [],
+            "skill_profile": [],
+            "family_profile": [],
+            "life_stage_profile": [],
+            "value_profile": [],
             "stable_traits": top_traits,
             "habit_patterns": top_habits,
             "long_term_reply_preferences": top_reply,
+            "group_role": "",
+            "decision_profile": "",
             "phase_state": top_topics[:3],
             "temperament_tendency": temperament,
             "summary_text": "本月沟通特征已按周摘要汇总。",
             "confidence": 0.5,
         }
 
+    def _format_group_messages_optimized(self, messages: List[Dict], member_name_map: Dict[str, str]) -> str:
+        if not messages:
+            return ""
+        time_groups = defaultdict(lambda: defaultdict(list))
+        for msg in messages:
+            timestamp = msg.get("timestamp")
+            sender = msg.get("sender", "")
+            content = str(msg.get("content", "")).strip()
+            if not sender or not content:
+                continue
+            try:
+                dt = datetime.strptime(str(timestamp), "%Y-%m-%d %H:%M:%S")
+            except Exception:
+                dt = None
+            if not dt:
+                continue
+            time_key = dt.strftime("%H:%M")
+            sender_name = member_name_map.get(sender, sender)
+            time_groups[time_key][sender_name].append(content)
+
+        lines = []
+        for time_key in sorted(time_groups.keys()):
+            lines.append(f"【{time_key}】")
+            for sender_name, contents in time_groups[time_key].items():
+                for idx, content in enumerate(contents):
+                    if idx == 0:
+                        lines.append(f"{sender_name}：{content}")
+                    else:
+                        lines.append(f"     {content}")
+        return "\n".join(lines)
+
     @staticmethod
     def _week_period_bounds(date_value: str) -> Tuple[str, str, str]:
         target_date = datetime.strptime(str(date_value)[:10], "%Y-%m-%d")

plugins/member_context/prompt_builder.py

@@ -6,6 +6,58 @@ from typing import Dict, List
 class MemberContextPromptBuilder:
     """成员分层画像提示词构建器"""
 
+    @staticmethod
+    def build_group_daily_digest_prompt(chatroom_id: str, digest_date: str,
+                                        member_labels: List[str], compressed_chat: str) -> str:
+        return (
+            "你是微信群后台的成员日观察批量提取器。\n"
+            "请基于给定的一天群聊记录，识别当天有参与发言的成员，并分别提取中性、克制的成员日观察摘要。\n"
+            "不要做心理诊断、隐私猜测、负面评价，不要输出未在候选名单中的成员。\n"
+            "你的输出将被后续系统按周、按月持续累积，因此优先提取可复用、可累计、可验证的行为信号，而不是一次性的情绪和玩笑。\n"
+            "输出严格 JSON，不要 markdown。\n"
+            "{"
+            "\"members\":["
+            "{"
+            "\"wxid\":\"成员wxid\","
+            "\"display_name\":\"成员显示名\","
+            "\"topics\":[\"主题1\"],"
+            "\"identity_clues\":[\"身份线索1\"],"
+            "\"skill_signals\":[\"技能信号1\"],"
+            "\"family_signals\":[\"家庭线索1\"],"
+            "\"life_stage_signals\":[\"阶段线索1\"],"
+            "\"value_preferences\":[\"价值偏好1\"],"
+            "\"interaction_style\":\"一句中文\","
+            "\"message_pattern\":\"一句中文\","
+            "\"response_style_hint\":\"一句中文\","
+            "\"habit_signals\":[\"信号1\"],"
+            "\"engagement_traits\":[\"特征1\"],"
+            "\"decision_style\":\"一句中文\","
+            "\"social_role\":\"一句中文，描述当天在群中的角色表现\","
+            "\"reply_taboos\":[\"避坑1\"],"
+            "\"temperament_signal\":\"一句中文，描述当天沟通倾向，必须克制\","
+            "\"summary_text\":\"一段不超过100字的日摘要\","
+            "\"representative_messages\":[\"原话1\",\"原话2\"],"
+            "\"confidence\":0.0"
+            "}"
+            "]"
+            "}\n"
+            "要求：\n"
+            "1. 只输出当天真正参与发言且能看出明确行为信号的成员；发言极少的人可以不输出。\n"
+            "2. 每个成员的 topics、identity_clues、skill_signals、family_signals、life_stage_signals、value_preferences、habit_signals、engagement_traits 最多4个，reply_taboos 最多3个。\n"
+            "3. representative_messages 只保留最能代表当天表达方式的短句，最多3条。\n"
+            "4. 必须严格使用候选成员列表中的 wxid 和显示名。\n"
+            "5. identity_clues、family_signals、life_stage_signals 只能写公开聊天中出现的线索，不可把弱线索写成确定事实。\n"
+            "6. skill_signals 重点提炼成员解决问题、提供信息、组织表达、专业能力等信号。\n"
+            "7. social_role 只描述当天在群里的角色表现，例如：问题提出者、信息补充者、气氛调节者、组织推进者。\n"
+            "8. topics 更偏向持续关注的话题方向；habit_signals 更偏向重复表达或互动习惯；engagement_traits 更偏向参与方式。\n"
+            "9. value_preferences 只记录公开表达出的偏好，如效率优先、成本敏感、谨慎验证、乐于助人，不要写抽象大词。\n"
+            "10. summary_text 应是后台观察摘要，不要写成对用户说的话。\n"
+            f"群ID: {chatroom_id}\n"
+            f"日期: {digest_date}\n"
+            "候选成员:\n" + "\n".join(member_labels[:80]) + "\n"
+            "压缩后的群聊记录:\n" + compressed_chat
+        )
+
     @staticmethod
     def build_daily_digest_prompt(chatroom_id: str, wxid: str, display_name: str,
                                   digest_date: str, messages: List[Dict]) -> str:
@@ -20,14 +72,22 @@ class MemberContextPromptBuilder:
             "你是微信群后台的成员日观察摘要生成器。\n"
             "请仅基于给定的当日公开聊天记录，提取对后续互动有帮助的中性行为观察。\n"
             "不要做人格诊断、隐私猜测、负面评价，不要脑补群外信息。\n"
+            "这些日观察会被后续系统按周、按月持续累积，所以应优先输出长期可验证的行为线索，而不是一次性情绪。\n"
             "输出严格 JSON，不要 markdown。\n"
             "{"
             "\"topics\":[\"主题1\"],"
+            "\"identity_clues\":[\"身份线索1\"],"
+            "\"skill_signals\":[\"技能信号1\"],"
+            "\"family_signals\":[\"家庭线索1\"],"
+            "\"life_stage_signals\":[\"阶段线索1\"],"
+            "\"value_preferences\":[\"价值偏好1\"],"
             "\"interaction_style\":\"一句中文\","
             "\"message_pattern\":\"一句中文\","
             "\"response_style_hint\":\"一句中文\","
             "\"habit_signals\":[\"信号1\"],"
             "\"engagement_traits\":[\"特征1\"],"
+            "\"decision_style\":\"一句中文\","
+            "\"social_role\":\"一句中文\","
             "\"reply_taboos\":[\"避坑1\"],"
             "\"temperament_signal\":\"一句中文，描述当天显露的沟通倾向，必须克制\","
             "\"summary_text\":\"一段不超过100字的日摘要\","
@@ -35,9 +95,14 @@ class MemberContextPromptBuilder:
             "\"confidence\":0.0"
             "}\n"
             "要求：\n"
-            "1. topics 最多4个，habit_signals 最多4个，engagement_traits 最多4个，reply_taboos 最多3个。\n"
+            "1. topics、identity_clues、skill_signals、family_signals、life_stage_signals、value_preferences、habit_signals、engagement_traits 最多4个，reply_taboos 最多3个。\n"
             "2. temperament_signal 只能写当日可观察到的沟通倾向，不可上升为长期性格判断。\n"
             "3. representative_messages 保留最能代表当天风格的短句，最多3条。\n"
+            "4. identity_clues、family_signals、life_stage_signals 只能写线索，不可写成确定事实。\n"
+            "5. skill_signals 重点描述专业能力、工具熟练度、信息组织能力、问题解决能力等当天显露出的信号。\n"
+            "6. topics 尽量写持续关注方向，避免写一次性插话；habit_signals 只写当天已明显出现的表达或互动习惯。\n"
+            "7. value_preferences 只保留公开表达出的判断偏好，如效率优先、先验证再决策、重成本、重稳定。\n"
+            "8. summary_text 要像后台备注，不要像对话回复。\n"
             f"成员: {display_name} ({wxid})\n"
             f"群ID: {chatroom_id}\n"
             f"日期: {digest_date}\n"
@@ -54,9 +119,16 @@ class MemberContextPromptBuilder:
                 "period_key": item.get("period_key"),
                 "summary_text": item.get("summary_text", ""),
                 "topics": structured.get("topics") or structured.get("stable_topics") or structured.get("long_term_topics") or [],
+                "identity_clues": structured.get("identity_clues") or structured.get("identity_traits") or [],
+                "skill_signals": structured.get("skill_signals") or structured.get("skill_profile") or [],
+                "family_signals": structured.get("family_signals") or structured.get("family_profile") or [],
+                "life_stage_signals": structured.get("life_stage_signals") or structured.get("life_stage_profile") or [],
+                "value_preferences": structured.get("value_preferences") or structured.get("value_profile") or [],
                 "habit_signals": structured.get("habit_signals") or structured.get("habit_patterns") or [],
                 "engagement_traits": structured.get("engagement_traits") or structured.get("stable_traits") or [],
                 "reply_preferences": structured.get("reply_preferences") or structured.get("long_term_reply_preferences") or [],
+                "social_role": structured.get("social_role") or structured.get("group_role") or "",
+                "decision_style": structured.get("decision_style") or structured.get("decision_profile") or "",
                 "temperament_signal": structured.get("temperament_signal") or structured.get("temperament_tendency") or "",
                 "recent_state": structured.get("recent_state") or [],
             }
@@ -66,9 +138,16 @@ class MemberContextPromptBuilder:
             schema = (
                 "{"
                 "\"stable_topics\":[\"主题1\"],"
+                "\"identity_traits\":[\"身份特征1\"],"
+                "\"skill_profile\":[\"技能画像1\"],"
+                "\"family_profile\":[\"家庭线索1\"],"
+                "\"life_stage_profile\":[\"阶段线索1\"],"
+                "\"value_profile\":[\"价值偏好1\"],"
                 "\"stable_traits\":[\"特征1\"],"
                 "\"habit_patterns\":[\"习惯1\"],"
                 "\"reply_preferences\":[\"偏好1\"],"
+                "\"group_role\":\"一句中文\","
+                "\"decision_profile\":\"一句中文\","
                 "\"recent_state\":[\"状态1\"],"
                 "\"temperament_tendency\":\"一句中文\","
                 "\"summary_text\":\"一段不超过120字的周摘要\","
@@ -80,9 +159,16 @@ class MemberContextPromptBuilder:
             schema = (
                 "{"
                 "\"long_term_topics\":[\"主题1\"],"
+                "\"identity_traits\":[\"身份特征1\"],"
+                "\"skill_profile\":[\"技能画像1\"],"
+                "\"family_profile\":[\"家庭线索1\"],"
+                "\"life_stage_profile\":[\"阶段线索1\"],"
+                "\"value_profile\":[\"价值偏好1\"],"
                 "\"stable_traits\":[\"特征1\"],"
                 "\"habit_patterns\":[\"习惯1\"],"
                 "\"long_term_reply_preferences\":[\"偏好1\"],"
+                "\"group_role\":\"一句中文\","
+                "\"decision_profile\":\"一句中文\","
                 "\"phase_state\":[\"状态1\"],"
                 "\"temperament_tendency\":\"一句中文\","
                 "\"summary_text\":\"一段不超过140字的月摘要\","
@@ -100,6 +186,10 @@ class MemberContextPromptBuilder:
             "1. 所有列表字段最多5项，必须中性克制。\n"
             "2. 只有多个下级摘要反复出现的特征，才允许写进 stable_traits / habit_patterns / long_term_reply_preferences。\n"
             "3. recent_state / phase_state 只描述当前阶段状态，不要冒充长期人格。\n"
+            "4. identity_traits、family_profile、life_stage_profile 只能保留反复出现的公开线索，不可编造事实。\n"
+            "5. skill_profile 要优先提炼稳定出现的能力、专业方向、擅长处理的问题类型。\n"
+            "6. group_role 描述其在群中的长期角色位置，decision_profile 描述其决策与判断风格。\n"
+            "7. value_profile 需要优先保留真正反复出现的判断偏好，如效率优先、成本敏感、风险谨慎、愿意分享。\n"
             f"成员: {display_name} ({wxid})\n"
             f"群ID: {chatroom_id}\n"
             f"周期: {period_key}\n"
@@ -125,9 +215,16 @@ class MemberContextPromptBuilder:
             "\"response_style_hint\":\"一句中文\","
             "\"topics_of_interest\":[\"主题1\"],"
             "\"recent_focus\":[\"近期主题1\"],"
+            "\"identity_traits\":[\"身份线索1\"],"
+            "\"skill_profile\":[\"技能画像1\"],"
+            "\"family_profile\":[\"家庭线索1\"],"
+            "\"life_stage_profile\":[\"阶段线索1\"],"
+            "\"value_profile\":[\"价值偏好1\"],"
             "\"stable_traits\":[\"长期特征1\"],"
             "\"habit_patterns\":[\"习惯1\"],"
             "\"long_term_reply_preferences\":[\"偏好1\"],"
+            "\"group_role\":\"一句中文\","
+            "\"decision_profile\":\"一句中文\","
             "\"recent_state\":[\"近期状态1\"],"
             "\"temperament_tendency\":\"一句中文\","
             "\"summary_text\":\"一段不超过150字的后台摘要\","
@@ -139,6 +236,10 @@ class MemberContextPromptBuilder:
             "1. stable_traits、habit_patterns、long_term_reply_preferences 只从月级和多次重复证据中提取。\n"
             "2. recent_focus、recent_state 更依赖最近周级和日级。\n"
             "3. summary_text 要像后台备注，不要明显暴露在给用户做画像。\n"
+            "4. identity_traits、family_profile、life_stage_profile 必须写成公开线索或长期观察，不得伪造事实。\n"
+            "5. skill_profile 要尽量覆盖专业能力、问题解决能力、表达组织能力、资源协调能力等维度。\n"
+            "6. group_role 要描述其在群中的角色定位，decision_profile 要描述其决策/判断方式。\n"
+            "7. 如果月级与周级证据不足，宁可少写，也不要把短期状态写成长期人格。\n"
             f"成员: {display_name} ({wxid})\n"
             f"群ID: {chatroom_id}\n"
             "月级摘要:\n" + ("\n".join(monthly_lines) or "暂无")

plugins/member_context/service.py

@@ -67,7 +67,8 @@ class MemberContextService:
         self.min_group_messages = int(schedule_config.get("min_group_messages", 20))
 
     def build_member_context(self, chatroom_id: str, wxid: str, days: Optional[int] = None,
-                             limit: Optional[int] = None, force_digest_rebuild: bool = False) -> Dict:
+                             limit: Optional[int] = None, force_digest_rebuild: bool = False,
+                             ensure_group_daily: bool = True) -> Dict:
         days = days or self.sample_days
         limit = limit or self.refresh_limit_per_member
 
@@ -75,6 +76,11 @@ class MemberContextService:
         member = self.contacts_db.get_chatroom_member_info(chatroom_id, wxid) or {}
         display_name = member.get("display_name") or member.get("nick_name") or wxid
 
+        group_digest_stats = {"built_daily": 0, "touched_members": []}
+        if ensure_group_daily:
+            group_digest_stats = self.digest_service.ensure_recent_group_daily_digests(
+                chatroom_id, force=force_digest_rebuild
+            )
         digest_snapshot = self.digest_service.ensure_member_digest_pipeline(
             chatroom_id, wxid, force=force_digest_rebuild
         )
@@ -82,7 +88,13 @@ class MemberContextService:
         weekly_digests = digest_snapshot.get("weekly_digests", [])
         monthly_digests = digest_snapshot.get("monthly_digests", [])
 
-        recent_messages = self.message_db.get_member_recent_messages(chatroom_id, wxid, days=min(days, 7), limit=120)
+        recent_messages = self.message_db.get_member_recent_messages(
+            chatroom_id,
+            wxid,
+            days=min(days, 7),
+            limit=120,
+            include_today=False,
+        )
         monthly_structured = [item.get("structured", {}) or {} for item in monthly_digests]
         weekly_structured = [item.get("structured", {}) or {} for item in weekly_digests]
         daily_structured = [item.get("structured", {}) or {} for item in daily_digests]
@@ -113,6 +125,26 @@ class MemberContextService:
             "source_days": days,
             "last_profiled_at": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
             "meta": {
+                "identity_traits": self._extract_scored_items(
+                    monthly_structured + weekly_structured + daily_structured,
+                    ["identity_traits", "identity_clues"], limit=5
+                ),
+                "skill_profile": self._extract_scored_items(
+                    monthly_structured + weekly_structured + daily_structured,
+                    ["skill_profile", "skill_signals"], limit=6
+                ),
+                "family_profile": self._extract_scored_items(
+                    monthly_structured + weekly_structured + daily_structured,
+                    ["family_profile", "family_signals"], limit=4
+                ),
+                "life_stage_profile": self._extract_scored_items(
+                    monthly_structured + weekly_structured + daily_structured,
+                    ["life_stage_profile", "life_stage_signals"], limit=4
+                ),
+                "value_profile": self._extract_scored_items(
+                    monthly_structured + weekly_structured + daily_structured,
+                    ["value_profile", "value_preferences"], limit=5
+                ),
                 "stable_traits": self._extract_scored_items(
                     monthly_structured + weekly_structured, ["stable_traits", "engagement_traits"], limit=self.stable_max_items
                 ),
@@ -130,6 +162,14 @@ class MemberContextService:
                     monthly_structured + weekly_structured + daily_structured,
                     ["temperament_tendency", "temperament_signal"], default=""
                 ),
+                "group_role": self._best_text(
+                    monthly_structured + weekly_structured + daily_structured,
+                    ["group_role", "social_role"], default=""
+                ),
+                "decision_profile": self._best_text(
+                    monthly_structured + weekly_structured + daily_structured,
+                    ["decision_profile", "decision_style"], default=""
+                ),
                 "engagement_traits": self._extract_scored_items(
                     daily_structured + weekly_structured, ["engagement_traits", "stable_traits"], limit=4
                 ),
@@ -144,7 +184,7 @@ class MemberContextService:
                 "last_daily_digest_at": daily_digests[0].get("last_generated_at") if daily_digests else "",
                 "last_weekly_digest_at": weekly_digests[0].get("last_generated_at") if weekly_digests else "",
                 "last_monthly_digest_at": monthly_digests[0].get("last_generated_at") if monthly_digests else "",
-                "refresh_mode": self._build_refresh_mode(existing_context, digest_snapshot),
+                "refresh_mode": self._build_refresh_mode(existing_context, digest_snapshot, group_digest_stats),
             },
         }
 
@@ -214,6 +254,13 @@ class MemberContextService:
             f"days={days}, limit_per_member={limit_per_member}"
         )
 
+        group_digest_stats = self.digest_service.ensure_recent_group_daily_digests(chatroom_id)
+        self.LOG.info(
+            f"[成员交互摘要] 群日摘要批处理完成: group={chatroom_id}, "
+            f"built_daily={group_digest_stats.get('built_daily', 0)}, "
+            f"touched_members={len(group_digest_stats.get('touched_members', []))}"
+        )
+
         for index, active_member in enumerate(active_members, start=1):
             wxid = active_member.get("wxid")
             if wxid not in enabled_members:
@@ -231,7 +278,9 @@ class MemberContextService:
                     f"last_profiled_at={(existing_context or {}).get('last_profiled_at')}"
                 )
                 continue
-            context = self.build_member_context(chatroom_id, wxid, days=days, limit=limit_per_member)
+            context = self.build_member_context(
+                chatroom_id, wxid, days=days, limit=limit_per_member, ensure_group_daily=False
+            )
             if context["source_message_count"] <= 0 and context.get("meta", {}).get("digest_daily_count", 0) <= 0:
                 skipped += 1
                 self.LOG.debug(
@@ -329,7 +378,8 @@ class MemberContextService:
         WHERE group_id = %s
           AND sender IS NOT NULL
           AND sender <> ''
-          AND timestamp >= DATE_SUB(NOW(), INTERVAL %s HOUR)
+          AND timestamp >= DATE_SUB(CURDATE(), INTERVAL %s HOUR)
+          AND timestamp < CURDATE()
           AND message_type IN (1, 49)
         GROUP BY sender
         HAVING COUNT(*) >= %s
@@ -384,7 +434,8 @@ class MemberContextService:
         SELECT group_id, COUNT(*) AS msg_count
         FROM messages
         WHERE group_id LIKE %s
-          AND timestamp >= DATE_SUB(NOW(), INTERVAL %s HOUR)
+          AND timestamp >= DATE_SUB(CURDATE(), INTERVAL %s HOUR)
+          AND timestamp < CURDATE()
         GROUP BY group_id
         HAVING COUNT(*) >= %s
         """
@@ -473,9 +524,16 @@ class MemberContextService:
             "summary_text": str(data.get("summary_text", "")).strip(),
             "confidence": max(0.0, min(1.0, confidence)),
             "meta": {
+                "identity_traits": norm_list(data.get("identity_traits"), 5),
+                "skill_profile": norm_list(data.get("skill_profile"), 6),
+                "family_profile": norm_list(data.get("family_profile"), 4),
+                "life_stage_profile": norm_list(data.get("life_stage_profile"), 4),
+                "value_profile": norm_list(data.get("value_profile"), 5),
                 "stable_traits": norm_list(data.get("stable_traits"), self.stable_max_items),
                 "habit_patterns": norm_list(data.get("habit_patterns"), self.stable_max_items),
                 "long_term_reply_preferences": norm_list(data.get("long_term_reply_preferences"), 4),
+                "group_role": str(data.get("group_role", "")).strip(),
+                "decision_profile": str(data.get("decision_profile", "")).strip(),
                 "recent_state": norm_list(data.get("recent_state"), 4),
                 "temperament_tendency": str(data.get("temperament_tendency", "")).strip(),
                 "engagement_traits": norm_list(data.get("engagement_traits"), 4),
@@ -510,6 +568,31 @@ class MemberContextService:
             meta.get("habit_patterns", []),
             current_context.get("confidence", 0),
         )
+        merged_identity_scores = self._merge_scored_items(
+            existing_meta.get("identity_trait_scores", {}),
+            meta.get("identity_traits", []),
+            current_context.get("confidence", 0) * 0.75,
+        )
+        merged_skill_scores = self._merge_scored_items(
+            existing_meta.get("skill_profile_scores", {}),
+            meta.get("skill_profile", []),
+            current_context.get("confidence", 0) * 0.85,
+        )
+        merged_family_scores = self._merge_scored_items(
+            existing_meta.get("family_profile_scores", {}),
+            meta.get("family_profile", []),
+            current_context.get("confidence", 0) * 0.55,
+        )
+        merged_life_stage_scores = self._merge_scored_items(
+            existing_meta.get("life_stage_profile_scores", {}),
+            meta.get("life_stage_profile", []),
+            current_context.get("confidence", 0) * 0.65,
+        )
+        merged_value_scores = self._merge_scored_items(
+            existing_meta.get("value_profile_scores", {}),
+            meta.get("value_profile", []),
+            current_context.get("confidence", 0) * 0.75,
+        )
         merged_reply_pref_scores = self._merge_scored_items(
             existing_meta.get("long_term_reply_preference_scores", {}),
             meta.get("long_term_reply_preferences", []),
@@ -524,13 +607,35 @@ class MemberContextService:
         meta["topic_scores"] = merged_topic_scores
         meta["stable_trait_scores"] = merged_trait_scores
         meta["habit_pattern_scores"] = merged_habit_scores
+        meta["identity_trait_scores"] = merged_identity_scores
+        meta["skill_profile_scores"] = merged_skill_scores
+        meta["family_profile_scores"] = merged_family_scores
+        meta["life_stage_profile_scores"] = merged_life_stage_scores
+        meta["value_profile_scores"] = merged_value_scores
         meta["long_term_reply_preference_scores"] = merged_reply_pref_scores
         meta["temperament_tendency_scores"] = merged_temperament_scores
+        meta["identity_traits"] = self._top_scored_items(merged_identity_scores, limit=5)
+        meta["skill_profile"] = self._top_scored_items(merged_skill_scores, limit=6)
+        meta["family_profile"] = self._top_scored_items(merged_family_scores, limit=4)
+        meta["life_stage_profile"] = self._top_scored_items(merged_life_stage_scores, limit=4)
+        meta["value_profile"] = self._top_scored_items(merged_value_scores, limit=5)
         meta["stable_traits"] = self._top_scored_items(merged_trait_scores, limit=self.stable_max_items)
         meta["habit_patterns"] = self._top_scored_items(merged_habit_scores, limit=self.stable_max_items)
         meta["long_term_reply_preferences"] = self._top_scored_items(merged_reply_pref_scores, limit=4)
         temperament = self._top_scored_items(merged_temperament_scores, limit=1)
         meta["temperament_tendency"] = temperament[0] if temperament else meta.get("temperament_tendency", "")
+        if not meta["identity_traits"]:
+            meta["identity_traits"] = (existing_meta.get("identity_traits") or [])[:5]
+        if not meta["skill_profile"]:
+            meta["skill_profile"] = (existing_meta.get("skill_profile") or [])[:6]
+        if not meta["family_profile"]:
+            meta["family_profile"] = (existing_meta.get("family_profile") or [])[:4]
+        if not meta["life_stage_profile"]:
+            meta["life_stage_profile"] = (existing_meta.get("life_stage_profile") or [])[:4]
+        if not meta["value_profile"]:
+            meta["value_profile"] = (existing_meta.get("value_profile") or [])[:5]
+        meta["group_role"] = meta.get("group_role") or existing_meta.get("group_role") or ""
+        meta["decision_profile"] = meta.get("decision_profile") or existing_meta.get("decision_profile") or ""
         meta["engagement_traits"] = (meta.get("engagement_traits") or existing_meta.get("engagement_traits") or [])[:4]
         meta["reply_taboos"] = (meta.get("reply_taboos") or existing_meta.get("reply_taboos") or [])[:3]
         meta["recent_state"] = (meta.get("recent_state") or existing_meta.get("recent_state") or [])[:4]
@@ -611,9 +716,12 @@ class MemberContextService:
     def _sum_digest_source_count(daily_digests: List[Dict]) -> int:
         return sum(int(item.get("source_count", 0)) for item in daily_digests)
 
-    def _build_refresh_mode(self, existing_context: Optional[Dict], digest_snapshot: Dict) -> str:
+    def _build_refresh_mode(self, existing_context: Optional[Dict], digest_snapshot: Dict,
+                            group_digest_stats: Optional[Dict] = None) -> str:
         if not existing_context:
             return "bootstrap"
+        if (group_digest_stats or {}).get("built_daily", 0) > 0:
+            return "daily_rollup"
         if (digest_snapshot.get("stats", {}) or {}).get("built_monthly", 0) > 0:
             return "recalibration"
         return "incremental"
@@ -626,6 +734,12 @@ class MemberContextService:
             parts.append(f"{label}：{meta.get('temperament_tendency')}")
         if meta.get("stable_traits"):
             parts.append(f"长期特征：{'、'.join(meta.get('stable_traits')[:3])}")
+        if meta.get("identity_traits"):
+            parts.append(f"身份线索：{'、'.join(meta.get('identity_traits')[:2])}")
+        if meta.get("skill_profile"):
+            parts.append(f"技能画像：{'、'.join(meta.get('skill_profile')[:3])}")
+        if meta.get("value_profile"):
+            parts.append(f"判断偏好：{'、'.join(meta.get('value_profile')[:2])}")
         if meta.get("habit_patterns"):
             parts.append(f"习惯模式：{'、'.join(meta.get('habit_patterns')[:3])}")
         if meta.get("recent_state"):