refactor: add rule engine, risk service, quote state machine, and replay tests

core/pydantic_ai_agent.py

@@ -22,6 +22,10 @@ from pydantic_ai.models.openai import OpenAIChatModel
 from pydantic_ai.providers.openai import OpenAIProvider
 from dotenv import load_dotenv
 from utils.metrics_tracker import emit as metrics_emit
+from utils.observability import emit_activity, build_trace_id
+from core.quote_state_machine import QuoteStateMachine
+from core.rules import Rule, RuleContext, RuleEngine, RuleResult
+from services.risk_service import RiskService
 
 load_dotenv()
 
@@ -209,16 +213,14 @@ class CustomerServiceAgent:
 
     @staticmethod
     def _activity_log(event: str, **kwargs):
-        safe = {}
-        for k, v in kwargs.items():
-            if isinstance(v, str):
-                safe[k] = v[:240]
-            else:
-                safe[k] = v
-        try:
-            logger.info(f"[ACTIVITY] event={event} data={json.dumps(safe, ensure_ascii=False)}")
-        except Exception:
-            logger.info(f"[ACTIVITY] event={event} data={safe}")
+        emit_activity(
+            logger,
+            event=event,
+            trace_id=str(kwargs.pop("trace_id", "")),
+            customer_id=str(kwargs.pop("customer_id", "")),
+            result=str(kwargs.pop("result", "ok")),
+            **kwargs,
+        )
 
     def __init__(self, skills_dir: str = "skills"):
         self.api_key = os.getenv("OPENAI_API_KEY")
@@ -231,6 +233,9 @@ class CustomerServiceAgent:
             self.batch_quote_delay_turns = max(0, int(os.getenv("BATCH_QUOTE_DELAY_TURNS", "1")))
         except Exception:
             self.batch_quote_delay_turns = 1
+        self.quote_state_machine = QuoteStateMachine(delay_turns=self.batch_quote_delay_turns)
+        self.risk_service = RiskService()
+        self._pre_rule_engine = self._build_pre_rule_engine()
 
         if not self.api_key:
             raise ValueError("请设置 OPENAI_API_KEY 环境变量")
@@ -430,7 +435,7 @@ class CustomerServiceAgent:
         收图阶段回复默认走 AI 改写，失败时回退到固定模板。
         """
         # 首张收图先承接“我看一下”，避免机械地立刻催“发完统一报价”。
-        if scene == "collect_ack" and len(state.pending_image_urls) <= 1:
+        if scene == "collect_ack" and len(state.pending_image_urls) == 1:
             first_ack = [
                 "收到了，我先看一下哈，稍等哈",
                 "这张我收到了，我先看下，等我一下哈",
@@ -1272,21 +1277,7 @@ class CustomerServiceAgent:
     @staticmethod
     def _refresh_quote_phase(state: ConversationState, phase_hint: str = ""):
         """统一维护收图报价状态机。"""
-        if phase_hint in {"idle", "collecting", "ready_to_quote", "waiting_result"}:
-            state.quote_phase = phase_hint
-            if phase_hint == "idle":
-                state.quote_ready_turns = 0
-            return
-        if not state.pending_image_urls:
-            state.quote_phase = "idle"
-            state.quote_ready_turns = 0
-            return
-        if state.quote_phase in {"ready_to_quote", "waiting_result"}:
-            return
-        if state.pending_image_urls and state.pending_requirements:
-            state.quote_phase = "collecting"
-            return
-        state.quote_phase = "collecting"
+        QuoteStateMachine().refresh(state, phase_hint=phase_hint)
 
     def _should_defer_batch_quote(self, state: ConversationState, mark_ready: bool = False) -> bool:
         """
@@ -1294,19 +1285,13 @@ class CustomerServiceAgent:
         - 首次进入 ready_to_quote 时按配置等待 N 轮
         - 等待轮次归零后，本轮即可报价
         """
-        if mark_ready and state.quote_phase != "ready_to_quote":
-            state.quote_phase = "ready_to_quote"
-            state.quote_ready_turns = max(0, int(self.batch_quote_delay_turns))
-        if state.quote_phase == "ready_to_quote" and state.quote_ready_turns > 0:
-            state.quote_ready_turns -= 1
-            return True
-        return False
+        self.quote_state_machine.delay_turns = max(0, int(self.batch_quote_delay_turns))
+        return self.quote_state_machine.should_defer_batch_quote(state, mark_ready=mark_ready)
 
     def _mark_quote_ready(self, state: ConversationState):
         """仅标记 ready 状态，不消费等待轮次。"""
-        if state.quote_phase != "ready_to_quote":
-            state.quote_phase = "ready_to_quote"
-            state.quote_ready_turns = max(0, int(self.batch_quote_delay_turns))
+        self.quote_state_machine.delay_turns = max(0, int(self.batch_quote_delay_turns))
+        self.quote_state_machine.mark_ready(state)
 
     def _build_reject_message(self, reason: str = "") -> str:
         templates = [
@@ -1779,10 +1764,163 @@ class CustomerServiceAgent:
         clean = msg.strip().rstrip("！!？?。.～~")
         return clean in self._COOLDOWN_PATTERNS
 
+    def _build_pre_rule_engine(self) -> RuleEngine:
+        return RuleEngine(
+            rules=[
+                Rule(
+                    name="meaningless_short_text",
+                    priority=10,
+                    predicate=self._rule_pred_meaningless_short_text,
+                    action=self._rule_act_meaningless_short_text,
+                ),
+                Rule(
+                    name="cooldown_silent",
+                    priority=20,
+                    predicate=self._rule_pred_cooldown_silent,
+                    action=self._rule_act_cooldown_silent,
+                ),
+                Rule(
+                    name="manual_risk_block",
+                    priority=30,
+                    predicate=self._rule_pred_manual_risk_block,
+                    action=self._rule_act_manual_risk_block,
+                ),
+                Rule(
+                    name="text_risk_block",
+                    priority=40,
+                    predicate=self._rule_pred_text_risk_block,
+                    action=self._rule_act_text_risk_block,
+                ),
+            ]
+        )
+
+    async def _rule_pred_meaningless_short_text(self, ctx: RuleContext) -> bool:
+        message: CustomerMessage = ctx.get("message")
+        return _is_meaningless_short_text(message.msg)
+
+    async def _rule_act_meaningless_short_text(self, ctx: RuleContext) -> RuleResult:
+        message: CustomerMessage = ctx.get("message")
+        state: ConversationState = ctx.get("state")
+        trace_id = ctx.get("trace_id", "")
+        ping = random.choice(("嗯咯", "嗯啦", "嗯", "哦"))
+        state.last_reply_at = datetime.now()
+        self._activity_log(
+            "agent_ping_reply",
+            trace_id=trace_id,
+            customer_id=message.from_id,
+            msg=message.msg,
+            reply=ping,
+        )
+        return RuleResult(
+            matched=True,
+            stop=True,
+            action="agent_ping_reply",
+            payload={"response": AgentResponse(reply=ping, should_reply=True, need_transfer=False)},
+        )
+
+    async def _rule_pred_cooldown_silent(self, ctx: RuleContext) -> bool:
+        message: CustomerMessage = ctx.get("message")
+        state: ConversationState = ctx.get("state")
+        return self._in_cooldown(state, message.msg)
+
+    async def _rule_act_cooldown_silent(self, ctx: RuleContext) -> RuleResult:
+        message: CustomerMessage = ctx.get("message")
+        state: ConversationState = ctx.get("state")
+        trace_id = ctx.get("trace_id", "")
+        elapsed = int((datetime.now() - state.last_reply_at).total_seconds()) if state.last_reply_at else 0
+        print(f"[Agent] 冷却期静默（距上次回复 {elapsed}s）：{message.msg!r}")
+        self._activity_log(
+            "agent_cooldown_silent",
+            trace_id=trace_id,
+            customer_id=message.from_id,
+            elapsed_s=elapsed,
+        )
+        return RuleResult(
+            matched=True,
+            stop=True,
+            action="agent_cooldown_silent",
+            payload={"response": AgentResponse(reply="", should_reply=False, need_transfer=False)},
+        )
+
+    async def _rule_pred_manual_risk_block(self, ctx: RuleContext) -> bool:
+        message: CustomerMessage = ctx.get("message")
+        decision = self.risk_service.check_manual_block(message.from_id)
+        ctx.set("manual_risk_decision", decision)
+        return decision.blocked
+
+    async def _rule_act_manual_risk_block(self, ctx: RuleContext) -> RuleResult:
+        message: CustomerMessage = ctx.get("message")
+        trace_id = ctx.get("trace_id", "")
+        decision = ctx.get("manual_risk_decision")
+        self._activity_log(
+            "agent_manual_risk_reject",
+            trace_id=trace_id,
+            customer_id=message.from_id,
+            risk=(decision.profile if decision else {}),
+        )
+        return RuleResult(
+            matched=True,
+            stop=True,
+            action="agent_manual_risk_reject",
+            payload={
+                "response": AgentResponse(
+                    reply="这边无法继续为你处理该类需求，给你转人工专员对接。",
+                    should_reply=True,
+                    need_transfer=True,
+                    transfer_msg=TRANSFER_MESSAGE,
+                )
+            },
+        )
+
+    async def _rule_pred_text_risk_block(self, ctx: RuleContext) -> bool:
+        message: CustomerMessage = ctx.get("message")
+        decision = await self.risk_service.check_text_block(
+            message.msg,
+            political_detector=self._is_political_inquiry,
+            map_detector=self._is_map_inquiry,
+        )
+        ctx.set("text_risk_decision", decision)
+        return decision.blocked
+
+    async def _rule_act_text_risk_block(self, ctx: RuleContext) -> RuleResult:
+        message: CustomerMessage = ctx.get("message")
+        state: ConversationState = ctx.get("state")
+        trace_id = ctx.get("trace_id", "")
+        decision = ctx.get("text_risk_decision")
+        state.pending_image_urls.clear()
+        state.pending_requirements.clear()
+        self._sync_pending_quote_state(message.from_id, state)
+
+        reject_text = self.risk_service.build_reject_text(decision.category if decision else "other")
+        reply = await self._rewrite_reply_with_ai(
+            message=message,
+            state=state,
+            reply=reject_text,
+            scene="risk_reject",
+        )
+        state.last_reply_at = datetime.now()
+        print(f"{self.C_REPLY}[REPLY->CUSTOMER]{self.C_RESET} {reply}")
+        self._activity_log(
+            "agent_risk_reject",
+            trace_id=trace_id,
+            customer_id=message.from_id,
+            risk_category=(decision.category if decision else "other"),
+            risk_source=(decision.source if decision else "unknown"),
+            reply=reply,
+        )
+        return RuleResult(
+            matched=True,
+            stop=True,
+            action="agent_risk_reject",
+            payload={"response": AgentResponse(reply=reply, should_reply=True, need_transfer=False)},
+        )
+
     async def process_message(self, message: CustomerMessage) -> AgentResponse:
         """处理客户消息并生成回复"""
+        trace_id = build_trace_id(message.acc_id, message.from_id, message.msg_id, message.msg[:64])
         self._activity_log(
             "agent_inbound",
+            trace_id=trace_id,
             acc_id=message.acc_id,
             customer_id=message.from_id,
             msg=message.msg,
@@ -1791,97 +1929,12 @@ class CustomerServiceAgent:
         metrics_emit("inbound_msg", customer_id=message.from_id, acc_id=message.acc_id)
         # 获取或创建对话状态
         state = self._get_conversation_state(message.from_id)
-
-        # 无意义短句承接：单独回一句口语，不进入复杂决策
-        if _is_meaningless_short_text(message.msg):
-            ping = random.choice(("嗯咯", "嗯啦", "嗯", "哦"))
-            state.last_reply_at = datetime.now()
-            self._activity_log("agent_ping_reply", customer_id=message.from_id, msg=message.msg, reply=ping)
-            return AgentResponse(reply=ping, should_reply=True, need_transfer=False)
-
-        # 冷却期检测：近期已回复 + 纯打招呼 → 静默
-        if self._in_cooldown(state, message.msg):
-            elapsed = int((datetime.now() - state.last_reply_at).total_seconds())
-            print(f"[Agent] 冷却期静默（距上次回复 {elapsed}s）：{message.msg!r}")
-            self._activity_log("agent_cooldown_silent", customer_id=message.from_id, elapsed_s=elapsed)
-            return AgentResponse(reply="", should_reply=False, need_transfer=False)
-
-        # 前置风控：客户文本一旦命中政治/敏感询问，直接拒绝，避免“发图我看看”类答非所问
-        try:
-            # 人工风控：标记为不接单的客户直接转人工
-            manual_risk = risk_db.evaluate_customer(message.from_id)
-            if bool(manual_risk.get("do_not_serve")):
-                self._activity_log(
-                    "agent_manual_risk_reject",
-                    customer_id=message.from_id,
-                    risk=manual_risk,
-                )
-                return AgentResponse(
-                    reply="这边无法继续为你处理该类需求，给你转人工专员对接。",
-                    should_reply=True,
-                    need_transfer=True,
-                    transfer_msg=TRANSFER_MESSAGE,
-                )
-
-            from utils.content_filter import should_block_customer_smart
-            risk_hit, risk_category, _risk_reason = await should_block_customer_smart(message.msg)
-            map_hit = self._is_map_inquiry(message.msg) or (risk_category == "map")
-            political_hit = self._is_political_inquiry(message.msg) or (risk_category == "political")
-            if risk_hit or political_hit or map_hit:
-                # 命中敏感询问时清空待报价队列，避免旧图残留污染后续会话
-                state.pending_image_urls.clear()
-                state.pending_requirements.clear()
-                self._sync_pending_quote_state(message.from_id, state)
-                reject_text = "地图这类不做哈，这边不接地图相关需求。"
-                if risk_category == "sexual":
-                    reject_text = "这类不做哈，涉黄擦边内容都不接。"
-                elif risk_category == "violent":
-                    reject_text = "这类不做哈，暴力血腥相关都不接。"
-                elif political_hit and not map_hit:
-                    reject_text = "这类不做哈，政治相关图片和人物都不接。"
-                reply = await self._rewrite_reply_with_ai(
-                    message=message,
-                    state=state,
-                    reply=reject_text,
-                    scene="risk_reject",
-                )
-                state.last_reply_at = datetime.now()
-                print(f"{self.C_REPLY}[REPLY->CUSTOMER]{self.C_RESET} {reply}")
-                self._activity_log(
-                    "agent_risk_reject",
-                    customer_id=message.from_id,
-                    map_hit=map_hit,
-                    political_hit=political_hit,
-                    risk_category=risk_category,
-                    reply=reply,
-                )
-                return AgentResponse(reply=reply, should_reply=True, need_transfer=False)
-        except Exception:
-            map_hit = self._is_map_inquiry(message.msg)
-            political_hit = self._is_political_inquiry(message.msg)
-            if political_hit or map_hit:
-                state.pending_image_urls.clear()
-                state.pending_requirements.clear()
-                self._sync_pending_quote_state(message.from_id, state)
-                reject_text = "地图这类不做哈，这边不接地图相关需求。"
-                if political_hit and not map_hit:
-                    reject_text = "这类不做哈，政治相关图片和人物都不接。"
-                reply = await self._rewrite_reply_with_ai(
-                    message=message,
-                    state=state,
-                    reply=reject_text,
-                    scene="risk_reject",
-                )
-                state.last_reply_at = datetime.now()
-                print(f"{self.C_REPLY}[REPLY->CUSTOMER]{self.C_RESET} {reply}")
-                self._activity_log(
-                    "agent_risk_reject",
-                    customer_id=message.from_id,
-                    map_hit=map_hit,
-                    political_hit=political_hit,
-                    reply=reply,
-                )
-                return AgentResponse(reply=reply, should_reply=True, need_transfer=False)
+        pre_ctx = RuleContext(data={"message": message, "state": state, "trace_id": trace_id})
+        pre_result = await self._pre_rule_engine.run(pre_ctx)
+        if pre_result.stop:
+            response = pre_result.payload.get("response")
+            if isinstance(response, AgentResponse):
+                return response
 
         # 检测售前/售后
         new_stage = self._detect_stage(message.msg)

core/quote_state_machine.py

@@ -0,0 +1,47 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from typing import Protocol
+
+
+class QuoteStateLike(Protocol):
+    pending_image_urls: list
+    pending_requirements: list
+    quote_phase: str
+    quote_ready_turns: int
+
+
+@dataclass
+class QuoteStateMachine:
+    delay_turns: int = 1
+
+    def refresh(self, state: QuoteStateLike, phase_hint: str = "") -> None:
+        if phase_hint in {"idle", "collecting", "ready_to_quote", "waiting_result"}:
+            state.quote_phase = phase_hint
+            if phase_hint == "idle":
+                state.quote_ready_turns = 0
+            return
+
+        if not state.pending_image_urls:
+            state.quote_phase = "idle"
+            state.quote_ready_turns = 0
+            return
+
+        if state.quote_phase in {"ready_to_quote", "waiting_result"}:
+            return
+
+        state.quote_phase = "collecting"
+
+    def mark_ready(self, state: QuoteStateLike) -> None:
+        if state.quote_phase != "ready_to_quote":
+            state.quote_phase = "ready_to_quote"
+            state.quote_ready_turns = max(0, int(self.delay_turns))
+
+    def should_defer_batch_quote(self, state: QuoteStateLike, mark_ready: bool = False) -> bool:
+        if mark_ready and state.quote_phase != "ready_to_quote":
+            self.mark_ready(state)
+
+        if state.quote_phase == "ready_to_quote" and state.quote_ready_turns > 0:
+            state.quote_ready_turns -= 1
+            return True
+        return False

core/rules/__init__.py

@@ -0,0 +1,3 @@
+from .engine import Rule, RuleContext, RuleEngine, RuleResult
+
+__all__ = ["Rule", "RuleContext", "RuleEngine", "RuleResult"]

core/rules/engine.py

@@ -0,0 +1,59 @@
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from typing import Any, Awaitable, Callable, Dict, List, Optional
+
+
+@dataclass
+class RuleContext:
+    data: Dict[str, Any] = field(default_factory=dict)
+
+    def get(self, key: str, default: Any = None) -> Any:
+        return self.data.get(key, default)
+
+    def set(self, key: str, value: Any) -> None:
+        self.data[key] = value
+
+
+@dataclass
+class RuleResult:
+    matched: bool = False
+    stop: bool = False
+    action: str = ""
+    payload: Dict[str, Any] = field(default_factory=dict)
+
+
+Predicate = Callable[[RuleContext], Awaitable[bool]]
+Action = Callable[[RuleContext], Awaitable[RuleResult]]
+
+
+@dataclass
+class Rule:
+    name: str
+    priority: int
+    predicate: Predicate
+    action: Action
+
+
+class RuleEngine:
+    """Priority-ordered async rule chain."""
+
+    def __init__(self, rules: Optional[List[Rule]] = None):
+        self._rules: List[Rule] = sorted(rules or [], key=lambda x: x.priority)
+
+    def add_rule(self, rule: Rule) -> None:
+        self._rules.append(rule)
+        self._rules.sort(key=lambda x: x.priority)
+
+    async def run(self, ctx: RuleContext) -> RuleResult:
+        for rule in self._rules:
+            if not await rule.predicate(ctx):
+                continue
+            result = await rule.action(ctx)
+            if not result.matched:
+                result.matched = True
+            if not result.action:
+                result.action = rule.name
+            if result.stop:
+                return result
+        return RuleResult(matched=False, stop=False, action="no_match")

core/websocket_client.py

@@ -11,6 +11,7 @@ from collections import deque
 from datetime import datetime
 from pathlib import Path
 from typing import Optional, Dict, Any, List
+from utils.observability import emit_activity, build_trace_id
 
 # ========== 转接分组映射 ==========
 def _get_transfer_group(acc_id: str) -> str:
@@ -134,16 +135,14 @@ class QingjianAPIClient:
 
     def _activity_log(self, event: str, **kwargs):
         """统一活动日志，便于按 event 检索完整链路。"""
-        safe = {}
-        for k, v in kwargs.items():
-            if isinstance(v, str):
-                safe[k] = v[:200]
-            else:
-                safe[k] = v
-        try:
-            logger.info(f"[ACTIVITY] event={event} data={json.dumps(safe, ensure_ascii=False)}")
-        except Exception:
-            logger.info(f"[ACTIVITY] event={event} data={safe}")
+        emit_activity(
+            logger,
+            event=event,
+            trace_id=str(kwargs.pop("trace_id", "")),
+            customer_id=str(kwargs.pop("customer_id", "")),
+            result=str(kwargs.pop("result", "ok")),
+            **kwargs,
+        )
 
 
     async def connect(self):
@@ -721,6 +720,8 @@ class QingjianAPIClient:
         try:
             msg_text = self.to_chinese(data.get('msg', ''))
             _cid  = data.get('from_id', '')
+            trace_id = build_trace_id(data.get("acc_id", ""), _cid, data.get("msg_id", ""), msg_text[:64])
+            data["_trace_id"] = trace_id
             _name = self.to_chinese(data.get('from_name', '') or data.get('cy_name', ''))
             _plat = data.get('acc_type', '')
             _shop_type = _get_shop_type(data.get('acc_id', ''), self.to_chinese(data.get('goods_name', '') or ''))
@@ -864,19 +865,32 @@ class QingjianAPIClient:
             logger.info("Agent 正在处理消息...")
             self._activity_log(
                 "agent_process_start",
+                trace_id=trace_id,
                 acc_id=data.get("acc_id", ""),
                 customer_id=data.get("from_id", ""),
                 msg=msg_text,
             )
             
             # 调用 Agent
+            _t0 = time.monotonic()
             response = await self.agent.process_message(customer_msg)
+            self._activity_log(
+                "agent_process_done",
+                trace_id=trace_id,
+                acc_id=data.get("acc_id", ""),
+                customer_id=data.get("from_id", ""),
+                result="ok",
+                latency_ms=int((time.monotonic() - _t0) * 1000),
+                should_reply=bool(response.should_reply),
+                need_transfer=bool(response.need_transfer),
+            )
             
             # 检查是否需要转接人工
             if response.need_transfer:
                 logger.info("Agent 决定转接人工")
                 self._activity_log(
                     "agent_transfer",
+                    trace_id=trace_id,
                     acc_id=data.get("acc_id", ""),
                     customer_id=data.get("from_id", ""),
                     transfer_msg=response.transfer_msg,
@@ -932,6 +946,7 @@ class QingjianAPIClient:
                     logger.info(f"Agent 回复: {response.reply}")
                     self._activity_log(
                         "agent_reply",
+                        trace_id=trace_id,
                         acc_id=data.get("acc_id", ""),
                         customer_id=data.get("from_id", ""),
                         reply=response.reply,
@@ -955,6 +970,7 @@ class QingjianAPIClient:
                 logger.info("Agent 决定不回复此消息")
                 self._activity_log(
                     "agent_no_reply",
+                    trace_id=trace_id,
                     acc_id=data.get("acc_id", ""),
                     customer_id=data.get("from_id", ""),
                 )
@@ -963,6 +979,7 @@ class QingjianAPIClient:
             logger.error(f"Agent 处理失败: {e}")
             self._activity_log(
                 "agent_process_error",
+                trace_id=data.get("_trace_id", ""),
                 acc_id=data.get("acc_id", ""),
                 customer_id=data.get("from_id", ""),
                 error=str(e),
@@ -1659,10 +1676,12 @@ class QingjianAPIClient:
             original_msg: 收到的原始消息字典
             reply_content: 回复内容（文本或本地文件路径/http地址）
         """
+        trace_id = original_msg.get("_trace_id", "")
         if not self.websocket:
             print(f"[{self.get_time()}] 错误: 未连接到服务器")
             self._activity_log(
                 "send_reply_skipped",
+                trace_id=trace_id,
                 reason="websocket_not_connected",
                 acc_id=original_msg.get("acc_id", ""),
                 customer_id=original_msg.get("from_id", ""),
@@ -1687,6 +1706,7 @@ class QingjianAPIClient:
                 )
                 self._activity_log(
                     "send_reply_throttled",
+                    trace_id=trace_id,
                     key=ckey,
                     cooldown_s=cooldown,
                     msg=str(reply_content),
@@ -1716,10 +1736,12 @@ class QingjianAPIClient:
         self._log_outbound_once(original_msg, str(reply_content))
         self._activity_log(
             "send_reply_attempt",
+            trace_id=trace_id,
             acc_id=shop_id,
             customer_id=customer_id,
             msg=str(reply_content),
         )
+        reply["_trace_id"] = trace_id
         await self.send_message(reply)
 
     def _colloquialize_outbound_reply(self, text: Any) -> Any:
@@ -1820,6 +1842,7 @@ class QingjianAPIClient:
                 print(f"[{self.get_time()}] 发送成功:\n{pretty}")
                 self._activity_log(
                     "send_message_success",
+                    trace_id=message.get("_trace_id", ""),
                     acc_id=message.get("acc_id", ""),
                     customer_id=message.get("from_id", ""),
                     msg_type=message.get("msg_type", 0),
@@ -1829,6 +1852,7 @@ class QingjianAPIClient:
                 print(f"[{self.get_time()}] 发送失败: {e}")
                 self._activity_log(
                     "send_message_error",
+                    trace_id=message.get("_trace_id", ""),
                     acc_id=message.get("acc_id", ""),
                     customer_id=message.get("from_id", ""),
                     error=str(e),
@@ -1837,6 +1861,7 @@ class QingjianAPIClient:
             print(f"[{self.get_time()}] 错误: 连接未打开")
             self._activity_log(
                 "send_message_skipped",
+                trace_id=message.get("_trace_id", ""),
                 reason="socket_not_open",
                 acc_id=message.get("acc_id", ""),
                 customer_id=message.get("from_id", ""),

services/risk_service.py

@@ -0,0 +1,78 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from typing import Any, Callable, Dict, Awaitable
+
+from db.customer_risk_db import risk_db
+from utils.content_filter import should_block_customer_smart
+
+
+@dataclass
+class RiskDecision:
+    blocked: bool
+    category: str = "none"
+    reason: str = ""
+    source: str = "none"  # manual/ai_filter/fallback
+    profile: Dict[str, Any] | None = None
+
+
+class RiskService:
+    def evaluate_customer(self, customer_id: str) -> Dict[str, Any]:
+        return risk_db.evaluate_customer(customer_id)
+
+    def check_manual_block(self, customer_id: str) -> RiskDecision:
+        profile = self.evaluate_customer(customer_id)
+        if bool(profile.get("do_not_serve")):
+            return RiskDecision(
+                blocked=True,
+                category="manual_block",
+                reason="do_not_serve",
+                source="manual",
+                profile=profile,
+            )
+        return RiskDecision(blocked=False, source="manual", profile=profile)
+
+    async def check_text_block(
+        self,
+        text: str,
+        *,
+        political_detector: Callable[[str], bool],
+        map_detector: Callable[[str], bool],
+    ) -> RiskDecision:
+        try:
+            hit, category, reason = await should_block_customer_smart(text)
+            map_hit = map_detector(text) or category == "map"
+            political_hit = political_detector(text) or category == "political"
+            if hit or map_hit or political_hit:
+                c = category
+                if map_hit:
+                    c = "map"
+                elif political_hit:
+                    c = "political"
+                return RiskDecision(
+                    blocked=True,
+                    category=c or "other",
+                    reason=reason or "sensitive_text",
+                    source="ai_filter",
+                )
+            return RiskDecision(blocked=False, category="none", source="ai_filter")
+        except Exception:
+            map_hit = map_detector(text)
+            political_hit = political_detector(text)
+            if map_hit:
+                return RiskDecision(blocked=True, category="map", reason="fallback_match", source="fallback")
+            if political_hit:
+                return RiskDecision(blocked=True, category="political", reason="fallback_match", source="fallback")
+            return RiskDecision(blocked=False, category="none", source="fallback")
+
+    @staticmethod
+    def build_reject_text(category: str) -> str:
+        if category == "map":
+            return "地图这类不做哈，这边不接地图相关需求。"
+        if category == "sexual":
+            return "这类不做哈，涉黄擦边内容都不接。"
+        if category == "violent":
+            return "这类不做哈，暴力血腥相关都不接。"
+        if category == "political":
+            return "这类不做哈，政治相关图片和人物都不接。"
+        return "这类不做哈，这边不接这类需求。"

tests/replay/test_golden_replay.py

@@ -0,0 +1,42 @@
+import unittest
+
+from core.quote_state_machine import QuoteStateMachine
+
+
+class _State:
+    def __init__(self):
+        self.pending_image_urls = []
+        self.pending_requirements = []
+        self.quote_phase = "idle"
+        self.quote_ready_turns = 0
+
+
+class GoldenReplayTests(unittest.TestCase):
+    def test_replay_collect_then_ready_then_quote(self):
+        sm = QuoteStateMachine(delay_turns=1)
+        st = _State()
+
+        replay = [
+            {"event": "image", "url": "a.jpg", "want_phase": "collecting"},
+            {"event": "image", "url": "b.jpg", "want_phase": "collecting"},
+            {"event": "finish", "want_phase": "ready_to_quote", "want_defer": True},
+            {"event": "progress", "want_phase": "ready_to_quote", "want_defer": False},
+        ]
+
+        for step in replay:
+            if step["event"] == "image":
+                st.pending_image_urls.append(step["url"])
+                sm.refresh(st)
+                self.assertEqual(st.quote_phase, step["want_phase"])
+            elif step["event"] == "finish":
+                deferred = sm.should_defer_batch_quote(st, mark_ready=True)
+                self.assertEqual(st.quote_phase, step["want_phase"])
+                self.assertEqual(deferred, step["want_defer"])
+            elif step["event"] == "progress":
+                deferred = sm.should_defer_batch_quote(st, mark_ready=False)
+                self.assertEqual(st.quote_phase, step["want_phase"])
+                self.assertEqual(deferred, step["want_defer"])
+
+
+if __name__ == "__main__":
+    unittest.main(verbosity=2)

tests/test_rule_engine.py

@@ -0,0 +1,57 @@
+import unittest
+
+from core.quote_state_machine import QuoteStateMachine
+from core.rules import Rule, RuleContext, RuleEngine, RuleResult
+from services.risk_service import RiskService
+
+
+class _StubState:
+    def __init__(self):
+        self.pending_image_urls = []
+        self.pending_requirements = []
+        self.quote_phase = "idle"
+        self.quote_ready_turns = 0
+
+
+class RuleEngineTests(unittest.IsolatedAsyncioTestCase):
+    async def test_rule_engine_priority_and_stop(self):
+        async def pred_true(_ctx):
+            return True
+
+        async def act_first(_ctx):
+            return RuleResult(matched=True, stop=True, action="first", payload={"x": 1})
+
+        async def act_second(_ctx):
+            return RuleResult(matched=True, stop=True, action="second", payload={"x": 2})
+
+        engine = RuleEngine(
+            [
+                Rule(name="r2", priority=20, predicate=pred_true, action=act_second),
+                Rule(name="r1", priority=10, predicate=pred_true, action=act_first),
+            ]
+        )
+        out = await engine.run(RuleContext())
+        self.assertTrue(out.matched)
+        self.assertEqual(out.action, "first")
+        self.assertEqual(out.payload["x"], 1)
+
+    def test_quote_state_machine_transitions(self):
+        sm = QuoteStateMachine(delay_turns=1)
+        st = _StubState()
+        st.pending_image_urls = ["u1"]
+        sm.refresh(st)
+        self.assertEqual(st.quote_phase, "collecting")
+        self.assertTrue(sm.should_defer_batch_quote(st, mark_ready=True))
+        self.assertEqual(st.quote_phase, "ready_to_quote")
+        self.assertEqual(st.quote_ready_turns, 0)
+        self.assertFalse(sm.should_defer_batch_quote(st, mark_ready=False))
+
+    def test_risk_service_reject_text(self):
+        svc = RiskService()
+        self.assertIn("地图", svc.build_reject_text("map"))
+        self.assertIn("政治", svc.build_reject_text("political"))
+        self.assertIn("涉黄", svc.build_reject_text("sexual"))
+
+
+if __name__ == "__main__":
+    unittest.main(verbosity=2)

utils/observability.py

@@ -0,0 +1,89 @@
+from __future__ import annotations
+
+import hashlib
+import json
+import logging
+import time
+from typing import Any, Dict, Optional
+
+
+def build_trace_id(*parts: str) -> str:
+    raw = "|".join(str(p or "") for p in parts)
+    digest = hashlib.md5(raw.encode("utf-8")).hexdigest()
+    return digest[:16]
+
+
+def emit_activity(
+    logger: logging.Logger,
+    *,
+    event: str,
+    trace_id: str = "",
+    customer_id: str = "",
+    result: str = "ok",
+    latency_ms: Optional[int] = None,
+    **fields: Any,
+) -> None:
+    payload: Dict[str, Any] = {
+        "trace_id": trace_id or "-",
+        "customer_id": customer_id or "-",
+        "event": event,
+        "result": result,
+    }
+    if latency_ms is not None:
+        payload["latency_ms"] = int(max(0, latency_ms))
+    for k, v in (fields or {}).items():
+        if isinstance(v, str):
+            payload[k] = v[:400]
+        else:
+            payload[k] = v
+    try:
+        logger.info(f"[ACTIVITY] {json.dumps(payload, ensure_ascii=False)}")
+    except Exception:
+        logger.info(f"[ACTIVITY] {payload}")
+
+
+class ActivityTimer:
+    def __init__(
+        self,
+        *,
+        logger: logging.Logger,
+        event: str,
+        trace_id: str = "",
+        customer_id: str = "",
+        **fields: Any,
+    ):
+        self.logger = logger
+        self.event = event
+        self.trace_id = trace_id
+        self.customer_id = customer_id
+        self.fields = fields
+        self.start = time.monotonic()
+
+    def ok(self, **fields: Any) -> None:
+        elapsed_ms = int((time.monotonic() - self.start) * 1000)
+        merged = dict(self.fields)
+        merged.update(fields)
+        emit_activity(
+            self.logger,
+            event=self.event,
+            trace_id=self.trace_id,
+            customer_id=self.customer_id,
+            result="ok",
+            latency_ms=elapsed_ms,
+            **merged,
+        )
+
+    def fail(self, error: str, **fields: Any) -> None:
+        elapsed_ms = int((time.monotonic() - self.start) * 1000)
+        merged = dict(self.fields)
+        merged.update(fields)
+        emit_activity(
+            self.logger,
+            event=self.event,
+            trace_id=self.trace_id,
+            customer_id=self.customer_id,
+            result="error",
+            latency_ms=elapsed_ms,
+            error=error,
+            **merged,
+        )