Jelajahi Sumber

feat(ai-q): 小程序逐题对话问卷页 + 画像展示页

Xiaogang Liao 3 bulan lalu
induk
melakukan
61bbcf72ff
57 mengubah file dengan 4017 tambahan dan 56 penghapusan
  1. 12 0
      cfc-frontend/pages.json
  2. 18 0
      cfc-frontend/pages/health-main/index.vue
  3. 100 0
      cfc-frontend/pages/health/ai-questionnaire-result.vue
  4. 286 0
      cfc-frontend/pages/health/ai-questionnaire.vue
  5. 8 0
      cfc-frontend/utils/api.js
  6. 26 0
      cfc-langgraph/.env.example
  7. 33 0
      cfc-langgraph/Dockerfile
  8. 45 0
      cfc-langgraph/README.md
  9. 25 0
      cfc-langgraph/app/agents/analysis_agent.py
  10. 35 0
      cfc-langgraph/app/agents/chat_agent.py
  11. 57 0
      cfc-langgraph/app/agents/intent_classifier.py
  12. 79 0
      cfc-langgraph/app/agents/multimodal_agent.py
  13. 186 0
      cfc-langgraph/app/agents/report_parse_agent.py
  14. 320 0
      cfc-langgraph/app/api/adapter.py
  15. 45 0
      cfc-langgraph/app/api/analyze.py
  16. 58 0
      cfc-langgraph/app/api/chat.py
  17. 38 0
      cfc-langgraph/app/api/logs.py
  18. 135 0
      cfc-langgraph/app/api/meal.py
  19. 42 0
      cfc-langgraph/app/api/recommend.py
  20. 4 23
      cfc-langgraph/app/api/report_parse.py
  21. 45 0
      cfc-langgraph/app/api/tongue.py
  22. 96 0
      cfc-langgraph/app/graphs/analysis_graph.py
  23. 178 0
      cfc-langgraph/app/graphs/chat_graph.py
  24. 322 0
      cfc-langgraph/app/graphs/health_butler_graph.py
  25. 179 0
      cfc-langgraph/app/graphs/health_coach_graph.py
  26. 79 0
      cfc-langgraph/app/graphs/recommend_graph.py
  27. 14 27
      cfc-langgraph/app/main.py
  28. 99 0
      cfc-langgraph/app/middleware.py
  29. 3 0
      cfc-langgraph/app/models/__init__.py
  30. 18 0
      cfc-langgraph/app/models/chat.py
  31. 18 0
      cfc-langgraph/app/models/common.py
  32. 48 0
      cfc-langgraph/app/models/meal.py
  33. 29 0
      cfc-langgraph/app/models/recommend.py
  34. 6 4
      cfc-langgraph/app/monitoring.py
  35. 166 0
      cfc-langgraph/app/parsers/pdf_parser.py
  36. 77 0
      cfc-langgraph/app/tasks/knowledge_sync.py
  37. 88 0
      cfc-langgraph/app/tools/java_client.py
  38. 58 0
      cfc-langgraph/app/tools/product_tools.py
  39. 63 0
      cfc-langgraph/app/tools/report_tools.py
  40. 54 0
      cfc-langgraph/deploy.sh
  41. 68 0
      cfc-langgraph/docker-compose.yml
  42. 97 0
      cfc-langgraph/docs/deployment.md
  43. 121 0
      cfc-langgraph/docs/operations.md
  44. 9 0
      cfc-langgraph/prometheus.yml
  45. 34 0
      cfc-langgraph/pyproject.toml
  46. 0 2
      cfc-langgraph/src/app.py
  47. 14 0
      cfc-langgraph/start.sh
  48. 16 0
      cfc-langgraph/tests/conftest.py
  49. 30 0
      cfc-langgraph/tests/test_analyze.py
  50. 151 0
      cfc-langgraph/tests/test_auto_tag.py
  51. 18 0
      cfc-langgraph/tests/test_chat.py
  52. 88 0
      cfc-langgraph/tests/test_graph.py
  53. 37 0
      cfc-langgraph/tests/test_intent.py
  54. 25 0
      cfc-langgraph/tests/test_memory.py
  55. 44 0
      cfc-langgraph/tests/test_recommend.py
  56. 61 0
      cfc-langgraph/tests/test_report_parse.py
  57. 12 0
      cfc-langgraph/tests/test_tongue.py

+ 12 - 0
cfc-frontend/pages.json

@@ -912,6 +912,18 @@
           "style": {
             "navigationBarTitleText": "报告详情"
           }
+        },
+        {
+          "path": "pages/health/ai-questionnaire",
+          "style": {
+            "navigationBarTitleText": "AI 健康问卷"
+          }
+        },
+        {
+          "path": "pages/health/ai-questionnaire-result",
+          "style": {
+            "navigationBarTitleText": "问卷画像报告"
+          }
         }
       ]
     },

+ 18 - 0
cfc-frontend/pages/health-main/index.vue

@@ -43,6 +43,17 @@
           <text class="self-test-arrow">›</text>
         </view>
       </view>
+      <!-- AI 健康问卷入口 -->
+      <view class="ai-q-entry" @click="goAiQuestionnaire">
+        <view class="ai-q-head">
+          <text class="ai-q-icon">🤖</text>
+          <view class="ai-q-info">
+            <text class="ai-q-title">AI 健康问卷</text>
+            <text class="ai-q-sub">智能问答,生成您的健康画像</text>
+          </view>
+          <text class="ai-q-arrow">›</text>
+        </view>
+      </view>
       <!-- 数据预览(示例) -->
       <view class="demo-section">
         <view class="demo-section-head">
@@ -360,6 +371,7 @@ export default {
       uni.navigateTo({ url: '/pages/health/report-list' + (report && report.id ? '?reportId=' + report.id : '') })
     },
     goSelfTest: function() { uni.navigateTo({ url: '/pages/health/survey-questionnaire?dim=all' }) },
+    goAiQuestionnaire: function() { uni.navigateTo({ url: '/pages/health/ai-questionnaire' }) },
     goPlanSummary: function() {
       var url = '/pages/health/health-plan-summary'
       if (this.selectedMemberId) url += '?memberId=' + this.selectedMemberId
@@ -478,6 +490,12 @@ export default {
 .self-test-info { flex: 1; }
 .self-test-title { font-size: 28rpx; font-weight: bold; color: #fff; display: block; }
 .self-test-sub { font-size: 22rpx; color: rgba(255,255,255,0.8); display: block; margin-top: 4rpx; }
+.ai-q-entry { margin: 0 24rpx 16rpx; background: linear-gradient(135deg, #F97316, #F59E0B); border-radius: 20rpx; padding: 24rpx 28rpx; }
+.ai-q-head { display: flex; align-items: center; }
+.ai-q-icon { font-size: 40rpx; margin-right: 16rpx; }
+.ai-q-info { flex: 1; }
+.ai-q-title { font-size: 28rpx; font-weight: bold; color: #fff; display: block; }
+.ai-q-sub { font-size: 22rpx; color: rgba(255,255,255,0.85); display: block; margin-top: 4rpx; }
 .self-test-arrow { font-size: 36rpx; color: rgba(255,255,255,0.7); }
 .demo-section { margin: 0 24rpx 16rpx; }
 .demo-section-head { margin-bottom: 8rpx; }

+ 100 - 0
cfc-frontend/pages/health/ai-questionnaire-result.vue

@@ -0,0 +1,100 @@
+<template>
+  <view class="container">
+    <view class="result-header">
+      <text class="result-title">AI 问卷画像报告</text>
+    </view>
+
+    <!-- 用户画像 -->
+    <view class="section">
+      <text class="section-title">用户画像</text>
+      <view class="profile-item" v-for="(p, i) in userProfile" :key="'up' + i">
+        <view class="profile-head">
+          <text class="profile-dim">{{ p.dimension }}</text>
+          <text class="profile-score">{{ p.score }} 分</text>
+        </view>
+        <view class="score-bar-bg">
+          <view class="score-bar-fill" :style="'width:' + clamp(p.score) + '%'"></view>
+        </view>
+        <text class="profile-desc">{{ p.description }}</text>
+        <view class="evidence-list" v-if="p.evidence && p.evidence.length > 0">
+          <text class="evidence-item" v-for="(ev, ei) in p.evidence" :key="'ev' + ei">· {{ ev }}</text>
+        </view>
+      </view>
+    </view>
+
+    <!-- 需求画像 -->
+    <view class="section" v-if="needProfile.length > 0">
+      <text class="section-title">需求画像</text>
+      <view class="profile-item" v-for="(n, i) in needProfile" :key="'np' + i">
+        <text class="profile-dim">{{ n.dimension }}</text>
+        <text class="profile-desc">{{ n.description }}</text>
+        <text class="profile-suggest" v-if="n.suggestion">建议:{{ n.suggestion }}</text>
+      </view>
+    </view>
+
+    <view class="footer-btns">
+      <button class="btn-recommend" @click="goRecommend">查看相关推荐</button>
+      <button class="btn-again" @click="goHome">完成</button>
+    </view>
+  </view>
+</template>
+
+<script>
+export default {
+  data() {
+    return {
+      sessionId: '',
+      userProfile: [],
+      needProfile: []
+    }
+  },
+  onLoad: function(options) {
+    this.sessionId = options.sessionId || ''
+    if (options.profile) {
+      try {
+        var profile = JSON.parse(decodeURIComponent(options.profile))
+        this.userProfile = profile.userProfile || profile.user_profile || []
+        this.needProfile = profile.needProfile || profile.need_profile || []
+      } catch (e) {
+        this.userProfile = []
+        this.needProfile = []
+      }
+    }
+  },
+  methods: {
+    clamp: function(score) {
+      var s = Number(score) || 0
+      if (s < 0) return 0
+      if (s > 100) return 100
+      return s
+    },
+    goRecommend: function() {
+      uni.navigateTo({ url: '/pages/diet/index?source=ai-questionnaire' })
+    },
+    goHome: function() {
+      uni.switchTab({ url: '/pages/body/index' })
+    }
+  }
+}
+</script>
+
+<style scoped>
+.container { min-height: 100vh; background: #f5f5f5; padding: 24rpx; }
+.result-header { padding: 20rpx 0 30rpx; }
+.result-title { font-size: 38rpx; font-weight: bold; color: #333; }
+.section { margin-bottom: 30rpx; }
+.section-title { font-size: 30rpx; font-weight: bold; color: #333; display: block; margin-bottom: 16rpx; }
+.profile-item { background: #fff; border-radius: 16rpx; padding: 28rpx; margin-bottom: 16rpx; }
+.profile-head { display: flex; justify-content: space-between; align-items: center; }
+.profile-dim { font-size: 28rpx; color: #333; font-weight: bold; }
+.profile-score { font-size: 26rpx; color: #F97316; }
+.score-bar-bg { height: 14rpx; background: #f0f0f0; border-radius: 7rpx; margin-top: 14rpx; overflow: hidden; }
+.score-bar-fill { height: 100%; background: #F97316; border-radius: 7rpx; }
+.profile-desc { font-size: 26rpx; color: #666; line-height: 1.6; margin-top: 14rpx; display: block; }
+.profile-suggest { font-size: 26rpx; color: #0EA5E9; margin-top: 10rpx; display: block; }
+.evidence-list { margin-top: 10rpx; }
+.evidence-item { font-size: 24rpx; color: #999; display: block; margin-top: 4rpx; }
+.footer-btns { margin-top: 20rpx; }
+.btn-recommend { background: #F97316; color: #fff; border-radius: 12rpx; }
+.btn-again { background: #fff; color: #666; border-radius: 12rpx; margin-top: 16rpx; }
+</style>

+ 286 - 0
cfc-frontend/pages/health/ai-questionnaire.vue

@@ -0,0 +1,286 @@
+<template>
+  <view class="container">
+    <!-- 场景选择 -->
+    <view v-if="!sceneId" class="pick-wrap">
+      <view class="pick-title">选择问卷场景</view>
+      <view class="scene-item" v-for="(s, i) in scenes" :key="'scene' + i" @click="chooseScene(s)">
+        <text class="scene-name">{{ s.sceneName }}</text>
+        <text class="scene-desc">{{ s.description }}</text>
+      </view>
+    </view>
+
+    <!-- 成员选择 -->
+    <view v-else-if="!memberId" class="pick-wrap">
+      <view class="pick-title">选择填写人</view>
+      <view class="scene-item" v-for="(m, i) in members" :key="'mem' + i" @click="chooseMember(m)">
+        <text class="scene-name">{{ m.nickname || m.name || '成员' }}</text>
+      </view>
+    </view>
+
+    <!-- 答题 -->
+    <view v-else class="qa-wrap">
+      <view class="qa-header">
+        <text class="qa-progress">已答 {{ answeredCount }} / {{ maxQuestions || '-' }} 题</text>
+        <text class="qa-finish" @click="onFinishEarly">结束并生成画像</text>
+      </view>
+
+      <view v-if="loading" class="qa-loading">
+        <text class="qa-loading-text">AI 正在思考下一题...</text>
+      </view>
+
+      <view v-else-if="currentQuestion" class="qa-card">
+        <text class="qa-text">{{ currentQuestion.text }}</text>
+
+        <!-- 单选/多选 -->
+        <view v-if="currentQuestion.type === 'single' || currentQuestion.type === 'multi'" class="qa-options">
+          <view class="qa-option" v-for="(opt, oi) in options" :key="'opt' + oi"
+                :class="{ selected: isSelected(opt.id) }" @click="toggleOption(opt)">
+            <text class="qa-option-label">{{ opt.label }}</text>
+          </view>
+        </view>
+
+        <!-- 量表 -->
+        <view v-if="currentQuestion.type === 'scale' && currentQuestion.scale" class="qa-scale">
+          <text class="qa-scale-label">{{ currentQuestion.scale.minLabel }}</text>
+          <slider :min="currentQuestion.scale.min" :max="currentQuestion.scale.max" :value="scaleValue"
+                  activeColor="#F97316" @change="onScaleChange" class="qa-slider" />
+          <text class="qa-scale-label">{{ currentQuestion.scale.maxLabel }}</text>
+        </view>
+
+        <!-- 文本 -->
+        <view v-if="currentQuestion.type === 'text'" class="qa-textarea-wrap">
+          <textarea v-model="textAnswer" class="qa-textarea" placeholder="请输入您的回答" />
+        </view>
+      </view>
+
+      <view v-if="!loading && currentQuestion" class="qa-submit">
+        <button class="btn-next" :disabled="!canSubmit" @click="onAnswer">回答并继续</button>
+      </view>
+    </view>
+  </view>
+</template>
+
+<script>
+import { aiQSceneList, aiQStart, aiQAnswer, aiQFinish, getFamilyMembers } from '../../utils/api.js'
+
+export default {
+  data() {
+    return {
+      scenes: [],
+      members: [],
+      sceneId: null,
+      sceneName: '',
+      memberId: null,
+      sessionId: null,
+      currentQuestion: null,
+      options: [],
+      selected: {},
+      scaleValue: 5,
+      textAnswer: '',
+      answeredCount: 0,
+      maxQuestions: 12,
+      loading: false
+    }
+  },
+  computed: {
+    canSubmit: function() {
+      var q = this.currentQuestion
+      if (!q) return false
+      if (q.type === 'text') return this.textAnswer && this.textAnswer.trim().length > 0
+      if (q.type === 'scale') return true
+      if (q.type === 'multi') {
+        for (var k in this.selected) { if (this.selected[k]) return true }
+        return false
+      }
+      return this.selected[q.id || 'k'] === true
+    }
+  },
+  onLoad: function() {
+    this.loadScenes()
+    this.loadMembers()
+  },
+  methods: {
+    async loadScenes() {
+      try {
+        var res = await aiQSceneList({ enabledOnly: true })
+        if (res.code === 200) {
+          this.scenes = (res.data && res.data.items) || res.data || []
+        } else {
+          uni.showToast({ title: res.message || '加载场景失败', icon: 'none' })
+        }
+      } catch (e) {
+        uni.showToast({ title: '加载场景失败', icon: 'none' })
+      }
+    },
+    loadMembers: function() {
+      var self = this
+      getFamilyMembers().then(function(res) {
+        if (res && res.code === 200 && res.data) {
+          var list = res.data instanceof Array ? res.data : []
+          self.members = list.filter(function(m) { return m.id })
+        }
+      }).catch(function() { self.members = [] })
+    },
+    chooseScene: function(s) {
+      this.sceneId = s.id
+      this.sceneName = s.sceneName
+      this.maxQuestions = s.maxQuestions || 12
+      this.startQuestionnaire()
+    },
+    chooseMember: function(m) {
+      this.memberId = m.id
+      this.startQuestionnaire()
+    },
+    async startQuestionnaire() {
+      this.loading = true
+      try {
+        var res = await aiQStart({ sceneId: this.sceneId, memberId: this.memberId })
+        if (res.code === 200 && res.data) {
+          this.sessionId = res.data.sessionId
+          this.currentQuestion = res.data.question
+          this.answeredCount = res.data.answeredCount || 0
+          this.resetAnswer()
+        } else {
+          uni.showToast({ title: res.message || '开始失败', icon: 'none' })
+        }
+      } catch (e) {
+        uni.showToast({ title: '开始失败', icon: 'none' })
+      } finally {
+        this.loading = false
+      }
+    },
+    isSelected: function(optId) {
+      return this.selected[optId] === true
+    },
+    toggleOption: function(opt) {
+      if (this.currentQuestion.type === 'single') {
+        var single = {}
+        single[opt.id] = true
+        this.selected = single
+      } else {
+        this.selected[opt.id] = !this.selected[opt.id]
+      }
+    },
+    onScaleChange: function(e) {
+      this.scaleValue = e.detail.value
+    },
+    resetAnswer: function() {
+      this.selected = {}
+      this.textAnswer = ''
+      this.scaleValue = 5
+      if (this.currentQuestion && this.currentQuestion.options) {
+        this.options = this.currentQuestion.options
+      } else {
+        this.options = []
+      }
+    },
+    buildAnswer: function() {
+      var q = this.currentQuestion
+      if (!q) return ''
+      if (q.type === 'text') return this.textAnswer.trim()
+      if (q.type === 'scale') return String(this.scaleValue)
+      var labels = []
+      if (q.type === 'single') {
+        for (var k in this.selected) {
+          if (this.selected[k]) {
+            var opt = this.options.find(function(o) { return o.id === k })
+            labels.push(opt ? opt.label : k)
+          }
+        }
+        return labels[0] || ''
+      }
+      var multiLabels = []
+      for (var k2 in this.selected) {
+        if (this.selected[k2]) {
+          var opt2 = this.options.find(function(o) { return o.id === k2 })
+          multiLabels.push(opt2 ? opt2.label : k2)
+        }
+      }
+      return multiLabels.join(',')
+    },
+    async onAnswer() {
+      var answer = this.buildAnswer()
+      if (!answer) return
+      this.loading = true
+      try {
+        var res = await aiQAnswer({ sessionId: this.sessionId, answer: answer })
+        if (res.code === 200 && res.data) {
+          if (res.data.action === 'finish' || res.data.finished) {
+            this.goResult(res.data.profile || {})
+          } else {
+            this.currentQuestion = res.data.question
+            this.answeredCount = res.data.answeredCount || 0
+            this.resetAnswer()
+          }
+        } else {
+          uni.showToast({ title: res.message || '提交失败', icon: 'none' })
+        }
+      } catch (e) {
+        uni.showToast({ title: '网络异常,请重试', icon: 'none' })
+      } finally {
+        this.loading = false
+      }
+    },
+    onFinishEarly: function() {
+      var self = this
+      uni.showModal({
+        title: '结束问卷',
+        content: '将基于当前回答生成画像,确定结束吗?',
+        success: function(r) {
+          if (r.confirm) self.finishQuestionnaire()
+        }
+      })
+    },
+    async finishQuestionnaire() {
+      this.loading = true
+      try {
+        var res = await aiQFinish({ sessionId: this.sessionId })
+        if (res.code === 200 && res.data) {
+          this.goResult(res.data)
+        } else {
+          uni.showToast({ title: res.message || '生成失败', icon: 'none' })
+        }
+      } catch (e) {
+        uni.showToast({ title: '网络异常,请重试', icon: 'none' })
+      } finally {
+        this.loading = false
+      }
+    },
+    goResult: function(profile) {
+      uni.redirectTo({
+        url: '/pages/health/ai-questionnaire-result?profile=' + encodeURIComponent(JSON.stringify(profile)) +
+             '&sessionId=' + (this.sessionId || '')
+      })
+    }
+  }
+}
+</script>
+
+<style scoped>
+.container { min-height: 100vh; background: #f5f5f5; padding: 24rpx; }
+.pick-wrap { padding: 40rpx 0; }
+.pick-title { font-size: 34rpx; font-weight: bold; color: #333; margin-bottom: 24rpx; }
+.scene-item { background: #fff; border-radius: 16rpx; padding: 28rpx; margin-bottom: 20rpx; }
+.scene-name { font-size: 30rpx; color: #333; font-weight: bold; display: block; }
+.scene-desc { font-size: 26rpx; color: #999; margin-top: 8rpx; display: block; }
+.qa-header { display: flex; justify-content: space-between; align-items: center; padding: 10rpx 4rpx 20rpx; }
+.qa-progress { font-size: 26rpx; color: #999; }
+.qa-finish { font-size: 26rpx; color: #F97316; }
+.qa-loading { text-align: center; padding: 120rpx 0; }
+.qa-loading-text { font-size: 28rpx; color: #999; }
+.qa-card { background: #fff; border-radius: 16rpx; padding: 32rpx; }
+.qa-text { font-size: 32rpx; color: #333; line-height: 1.6; display: block; }
+.qa-option { padding: 22rpx; border: 2rpx solid #e0e0e0; border-radius: 12rpx; margin-top: 20rpx;
+  background: #fafafa; }
+.qa-option.selected { border-color: #F97316; background: #FFF7ED; }
+.qa-option-label { font-size: 28rpx; color: #333; }
+.qa-scale { display: flex; align-items: center; margin-top: 30rpx; }
+.qa-scale-label { font-size: 24rpx; color: #666; width: 80rpx; }
+.qa-slider { flex: 1; margin: 0 12rpx; }
+.qa-textarea-wrap { margin-top: 24rpx; }
+.qa-textarea { width: 100%; height: 200rpx; background: #fafafa; border: 2rpx solid #e0e0e0;
+  border-radius: 12rpx; padding: 16rpx; font-size: 28rpx; box-sizing: border-box; }
+.qa-submit { margin-top: 30rpx; }
+.btn-next { background: #F97316; color: #fff; border-radius: 12rpx; }
+.btn-next[disabled] { background: #ccc; }
+</style>

+ 8 - 0
cfc-frontend/utils/api.js

@@ -2656,3 +2656,11 @@ export const getFoodCautionList = () => request('/api/food/recommendation/cautio
 export const getFamilyPlatformBalance = () => request('/api/family-platform/balance', 'POST', {})
 export const getFamilyPlatformLogs = (data) => request('/api/family-platform/logs', 'POST', data)
 export const exchangeCouponByCf = (data) => request('/api/family-platform/exchange-coupon', 'POST', data)
+
+// ── AI 动态问卷 ──
+export const aiQSceneList = (data) => request('/api/ai-questionnaire/scene/list', 'POST', data)
+export const aiQStart = (data) => request('/api/ai-questionnaire/start', 'POST', data)
+export const aiQAnswer = (data) => request('/api/ai-questionnaire/answer', 'POST', data)
+export const aiQFinish = (data) => request('/api/ai-questionnaire/finish', 'POST', data)
+export const aiQHistory = (data) => request('/api/ai-questionnaire/history', 'POST', data)
+export const aiQProfileDetail = (data) => request('/api/ai-questionnaire/profile/detail', 'POST', data)

+ 26 - 0
cfc-langgraph/.env.example

@@ -0,0 +1,26 @@
+# LLM
+LLM_API_KEY=sk-your-key
+LLM_BASE_URL=https://api.deepseek.com/v1
+LLM_MODEL=deepseek-chat
+
+# Embedding
+EMBEDDING_API_KEY=${LLM_API_KEY}
+EMBEDDING_BASE_URL=https://api.deepseek.com/v1
+EMBEDDING_MODEL=text-embedding-v3
+
+# Java Backend
+JAVA_BASE_URL=http://localhost:9082
+JAVA_CONTEXT_URL=${JAVA_BASE_URL}/api/ai/context
+
+# LangSmith (optional)
+LANGCHAIN_TRACING_V2=false
+LANGCHAIN_API_KEY=
+LANGCHAIN_PROJECT=cfc-langgraph
+
+# Service
+SERVICE_HOST=0.0.0.0
+SERVICE_PORT=9000
+LOG_LEVEL=info
+
+# Chroma
+CHROMA_DB_PATH=./data/chroma_db

+ 33 - 0
cfc-langgraph/Dockerfile

@@ -0,0 +1,33 @@
+# ── 运行镜像 ──
+FROM python:3.12-slim
+
+ENV TZ=Asia/Shanghai
+RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
+
+WORKDIR /app
+
+COPY .venv-packages.tar.gz /tmp/pkgs.tar.gz
+RUN tar xzf /tmp/pkgs.tar.gz -C /usr/local/lib/python3.12/site-packages/ && \
+    rm /tmp/pkgs.tar.gz
+
+COPY app/ app/
+COPY src/ src/
+COPY data/ data/
+
+RUN useradd -m -u 1000 appuser && \
+    chown -R appuser:appuser /app && \
+    mkdir -p /data/chroma_db && \
+    chown -R appuser:appuser /data
+
+USER appuser
+
+EXPOSE 9000
+
+CMD ["python3", "-m", "gunicorn", "app.main:app", \
+     "--worker-class", "uvicorn.workers.UvicornWorker", \
+     "--bind", "0.0.0.0:9000", \
+     "--workers", "2", \
+     "--timeout", "300", \
+     "--keep-alive", "10", \
+     "--access-logfile", "-", \
+     "--error-logfile", "-"]

+ 45 - 0
cfc-langgraph/README.md

@@ -0,0 +1,45 @@
+# CFC LangGraph 问卷生成服务
+
+关系评估问卷 AI 出题 graph,部署到 `ai.etotem.com.cn`。
+
+## 启动
+
+```bash
+cd cfc-langgraph
+pip install -r requirements.txt
+cp .env.example .env  # 填写 OPENAI_API_KEY / OPENAI_BASE_URL / LLM_MODEL
+uvicorn src.app:app --host 0.0.0.0 --port 9000
+```
+
+## 端点
+
+| Method | Path | 说明 |
+|--------|------|------|
+| GET | `/health` | 健康检查 |
+| POST | `/api/v1/questionnaire/generate` | 生成关系问卷 JSON |
+
+## 测试
+
+```bash
+cd cfc-langgraph
+pytest tests/ -v
+```
+
+## 契约
+
+Request:
+```json
+{
+  "member_name": "小明",
+  "relationship_type": "child",
+  "family_context": {}
+}
+```
+
+Response:
+```json
+{
+  "questionnaire_json": "<stringified JSON>",
+  "version": "20260804115900"
+}
+```

+ 25 - 0
cfc-langgraph/app/agents/analysis_agent.py

@@ -0,0 +1,25 @@
+from app.tools.java_client import JavaClient
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class AnalysisAgent:
+    """报告解读 Agent: 拉取报告数据 + 问卷数据 + LLM 分析"""
+
+    def __init__(self):
+        self.java = JavaClient()
+
+    async def get_report_summary(self, report_id: int) -> dict:
+        """获取报告概要 (用于 LLM 上下文)"""
+        try:
+            client = await self.java._get_client()
+            resp = await client.post("/api/health/report/summary", json={
+                "reportId": report_id,
+            })
+            data = resp.json()
+            if data.get("code") == 200:
+                return data.get("data", {})
+        except Exception as e:
+            logger.warning("获取报告概要失败: %s", e)
+        return {}

+ 35 - 0
cfc-langgraph/app/agents/chat_agent.py

@@ -0,0 +1,35 @@
+from typing import Optional
+from app.tools.java_client import JavaClient
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class ChatAgent:
+    """家庭聊天 Agent: 包装业务逻辑, 被 StateGraph 节点调用"""
+
+    def __init__(self):
+        self.java = JavaClient()
+
+    async def load_context(self, user_id: int, child_id: Optional[int] = None) -> dict:
+        """从 Java 侧加载用户上下文"""
+        params = {}
+        if child_id:
+            params["childId"] = child_id
+        return await self.java.get_user_context(user_id, params)
+
+    async def extract_tasks(self, answer: str, user_id: int, conv_id: str) -> list[dict]:
+        """从 AI 回答中解析 [TASK] 标记"""
+        import re
+        tasks = []
+        pattern = r'\[TASK:\s*\{([^}]+)\}\]'
+        matches = re.findall(pattern, answer)
+        for match in matches:
+            task_info = {}
+            for kv in match.split(","):
+                if ":" in kv:
+                    k, v = kv.split(":", 1)
+                    task_info[k.strip()] = v.strip().strip('"').strip("'")
+            if task_info.get("title"):
+                tasks.append(task_info)
+        return tasks

+ 57 - 0
cfc-langgraph/app/agents/intent_classifier.py

@@ -0,0 +1,57 @@
+from enum import Enum
+from langchain_openai import ChatOpenAI
+from app.config import settings
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class Intent(str, Enum):
+    CHAT = "chat"           # 日常聊天 / 育儿交流
+    ANALYSIS = "analysis"   # 报告解读 / 数据分析
+    RECOMMEND = "recommend" # 商品 / 活动推荐
+    TASK = "task"           # 任务创建 / 进度查询
+    MIND = "mind"           # 情绪疏导 / 心理支持
+    HEALTH = "health"       # 健康咨询 / 舌诊
+
+
+CLASSIFY_PROMPT = """从用户消息中识别意图, 只返回意图代码, 不要解释:
+
+- chat: 日常聊天、育儿交流、询问建议(非具体商品/报告)
+- analysis: 报告解读\数据分析\趋势查看(提到"报告"/"分析"/"评分")
+- recommend: 商品\活动\文章推荐(提到"推荐"/"买"/"吃什么"/"适合")
+- task: 任务相关(提到"任务"/"打卡"/"完成"/"奖励")
+- mind: 情绪问题\心理支持(提到"心情"/"难过"/"焦虑"/"不开心")
+- health: 健康咨询\舌诊(提到"舌"/"健康"/"体质")
+
+用户消息: {query}
+
+历史上下文: {context}
+
+意图代码:
+"""
+
+
+class IntentClassifier:
+    def __init__(self):
+        self.llm = ChatOpenAI(
+            model=settings.llm_model,
+            api_key=settings.llm_api_key,
+            base_url=settings.llm_base_url,
+            temperature=settings.llm_temperature,
+            max_tokens=20,
+        )
+
+    async def classify(self, query: str, context: str = "") -> Intent:
+        prompt = CLASSIFY_PROMPT.format(query=query[:200], context=context[:300])
+        try:
+            resp = await self.llm.ainvoke(prompt)
+            intent_str = resp.content.strip().lower()
+            for intent in Intent:
+                if intent.value in intent_str:
+                    return intent
+            logger.warning("无法识别的意图: %s, 默认 chat", intent_str)
+            return Intent.CHAT
+        except Exception as e:
+            logger.warning("意图分类失败: %s, 默认 chat", e)
+            return Intent.CHAT

+ 79 - 0
cfc-langgraph/app/agents/multimodal_agent.py

@@ -0,0 +1,79 @@
+import httpx
+from typing import Optional
+from app.config import settings
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class TongueDiagnosisAgent:
+    """舌诊分析 Agent
+
+    当前实现: 代理到 Dify Workflow (多模态最成熟)
+    后续可替换: 直接调用多模态 LLM API
+    """
+
+    def __init__(self):
+        self.dify_base = settings.dify_base_url or ""
+        self.dify_api_key = settings.dify_tongue_api_key or ""
+
+    async def diagnose(
+        self,
+        image_url: str,
+        user_id: int,
+        additional_context: Optional[dict] = None,
+    ) -> dict:
+        """舌诊分析: 调用 Dify Workflow 或直接 LLM"""
+        if self.dify_base and self.dify_api_key:
+            return await self._via_dify(image_url, user_id, additional_context)
+        else:
+            return await self._via_llm(image_url)
+
+    async def _via_dify(
+        self, image_url: str, user_id: int, context: Optional[dict]
+    ) -> dict:
+        """通过 Dify Workflow 执行舌诊"""
+        url = f"{self.dify_base}/workflows/run"
+        headers = {
+            "Authorization": f"Bearer {self.dify_api_key}",
+            "Content-Type": "application/json",
+        }
+        inputs = {"tongue_image": {"type": "image", "url": image_url}}
+        if context:
+            inputs.update(context)
+
+        body = {
+            "inputs": inputs,
+            "user": str(user_id),
+            "response_mode": "blocking",
+        }
+
+        try:
+            async with httpx.AsyncClient(timeout=30) as client:
+                resp = await client.post(url, json=body, headers=headers)
+                data = resp.json()
+                if "data" in data and "outputs" in data["data"]:
+                    return data["data"]["outputs"]
+        except Exception as e:
+            logger.warning("Dify 舌诊失败: %s", e)
+
+        return self._mock_result()
+
+    async def _via_llm(self, image_url: str) -> dict:
+        """直接调用多模态 LLM (预留)"""
+        logger.warning("多模态 LLM 未配置, 返回模拟数据")
+        return self._mock_result()
+
+    def _mock_result(self) -> dict:
+        return {
+            "overall_assessment": "舌象基本正常, 舌质淡红, 苔薄白, 提示脾胃功能尚可。",
+            "indicators": [
+                {"code": "tongue_color", "value": "淡红"},
+                {"code": "coating_color", "value": "薄白"},
+                {"code": "coating_texture", "value": "润"},
+                {"code": "fissure", "value": "无"},
+                {"code": "teeth_mark", "value": "轻"},
+                {"code": "sublingual_vein", "value": "正常"},
+                {"code": "constitution", "value": "平和质"},
+            ],
+        }

+ 186 - 0
cfc-langgraph/app/agents/report_parse_agent.py

@@ -0,0 +1,186 @@
+"""
+报告解析 Agent:算法解析 + LLM 兜底
+"""
+import json
+import logging
+from typing import Optional
+
+from app.config import settings
+from app.parsers.pdf_parser import parse_report_pdf_with_fallback
+
+logger = logging.getLogger(__name__)
+
+# 算法解析器使用中文键名(与 extract_full_report_v5.py 一致),
+# Java 消费端期望英文键名;在此做双向映射,Java 读取英文键即可。
+_CN_TO_EN_OVERVIEW: dict[str, str] = {
+    '健康总分': 'overallScore',
+    '菌群健康': 'gutHealthScore',
+    '慢病控制': 'chronicDiseaseScore',
+    '营养均衡': 'nutritionScore',
+    '肠道菌群平衡': 'balanceScore',
+    '菌群多样性': 'diversityScore',
+    '有益菌': 'beneficialScore',
+    '有害菌': 'harmfulScore',
+    '核心菌属': 'coreGenusScore',
+}
+
+
+class ReportParseAgent:
+    """报告解析 Agent"""
+
+    def __init__(self):
+        self.llm_api_key = getattr(settings, 'llm_api_key', '')
+
+    @staticmethod
+    def _normalize_overview_keys(result: dict) -> None:
+        """将 overview 中的中文键名映射为英文键名(双向写入),兼容 Java 消费端。"""
+        overview = result.get('overview')
+        if not overview:
+            return
+        for _cn, _en in _CN_TO_EN_OVERVIEW.items():
+            if _cn in overview and _en not in overview:
+                overview[_en] = overview[_cn]
+
+    async def parse(self, file_path: str) -> dict:
+        """解析 PDF 报告,算法解析 + LLM 兜底"""
+        # 1. 算法解析
+        result = parse_report_pdf_with_fallback(file_path)
+        logger.info("算法解析完成: format=%s, overview_keys=%d",
+                     result.get('format'), len(result.get('overview', {})))
+
+        # 1.5 归一化 overview 键名:中文 → 英文(Java 消费端兼容)
+        self._normalize_overview_keys(result)
+
+        # 2. 如果解析不完整,LLM 兜底
+        if result.get('_parse_incomplete') or not result.get('disease_risks'):
+            logger.info("算法解析不完整,尝试 LLM 兜底")
+            llm_result = await self._parse_with_llm(file_path)
+            if llm_result:
+                # 合并 LLM 结果到算法结果上(LLM 覆盖缺失字段)
+                for key in ['disease_risks', 'nutrition', 'amino_acids',
+                            'vitamins', 'trace_elements', 'indicators']:
+                    if key in llm_result and not result.get(key):
+                        result[key] = llm_result[key]
+                if llm_result.get('overview'):
+                    for k, v in llm_result['overview'].items():
+                        if k not in result.get('overview', {}):
+                            result.setdefault('overview', {})[k] = v
+
+        # 清理内部标记
+        result.pop('_parse_incomplete', None)
+        return result
+
+    async def parse_generic(self, file_path: str, extra_context: Optional[dict] = None) -> dict:
+        """通用报告 LLM 解析(不经过算法解析,直接走 LLM)"""
+        try:
+            from PyPDF2 import PdfReader
+            reader = PdfReader(file_path)
+            text = '\n'.join(page.extract_text() or '' for page in reader.pages)
+
+            ctx_str = ""
+            if extra_context:
+                ctx_str = f"\n额外上下文:{json.dumps(extra_context, ensure_ascii=False)}"
+
+            prompt = f"""你是一个通用报告解析专家。请从以下PDF文本中提取结构化数据,返回JSON格式。
+
+报告文本内容:
+{text[:12000]}{ctx_str}
+
+请分析这份报告,推断它的类型和内容,然后按以下JSON Schema返回:
+{{
+    "reportType": "推断的报告类型名称",
+    "reportTypeFamily": "报告家族分类(如: dan/cognitive/gut_flora/health_check/other)",
+    "confidence": "high/medium/low",
+    "summary": {{
+        "personName": "姓名",
+        "reportDate": "报告日期",
+        "reportNumber": "报告编号",
+        "overallScore": "总分(如果有)",
+        "interpretation": "报告整体解读摘要"
+    }},
+    "indicators": [
+        {{"name": "指标名称", "value": "数值", "category": "分类", "status": "状态"}}
+    ],
+    "sections": [
+        {{"title": "段落标题", "content": "段落内容摘要", "items": [{{"name": "...", "value": "..."}}]}}
+    ],
+    "textFeatures": ["文本特征1", "文本特征2", ...]
+}}
+
+只返回JSON,不要其他文字。"""
+
+            if self.llm_api_key:
+                import httpx
+                async with httpx.AsyncClient(timeout=120) as client:
+                    resp = await client.post(
+                        f"{settings.llm_base_url}/chat/completions",
+                        json={
+                            "model": settings.llm_model or "gpt-4o",
+                            "messages": [{"role": "user", "content": prompt}],
+                            "temperature": 0.1,
+                        },
+                        headers={"Authorization": f"Bearer {self.llm_api_key}"},
+                    )
+                    resp.raise_for_status()
+                    data = resp.json()
+                    content = data['choices'][0]['message']['content']
+                    content = content.replace('```json', '').replace('```', '').strip()
+                    return json.loads(content)
+            else:
+                logger.warning("LLM 未配置,返回空")
+                return {"reportType": "unknown", "summary": {}, "indicators": [], "sections": []}
+        except Exception as e:
+            logger.error("通用LLM解析失败: %s", e)
+            return {"reportType": "unknown", "error": str(e), "summary": {}, "indicators": [], "sections": []}
+
+    async def _parse_with_llm(self, file_path: str) -> Optional[dict]:
+        """LLM 兜底解析"""
+        try:
+            from PyPDF2 import PdfReader
+            reader = PdfReader(file_path)
+            text = '\n'.join(page.extract_text() or '' for page in reader.pages)
+
+            # 构造 prompt
+            prompt = f"""你是一个肠道菌群检测报告解析专家。请从以下PDF文本中提取结构化数据,返回JSON格式。
+
+文本内容:
+{text[:8000]}
+
+请按以下JSON Schema返回:
+{{
+    "overview": {{ "person_name": "", "report_number": "", "age": 0, "gender": "male/female",
+        "overallScore": 0, "gutHealthScore": 0, "chronicDiseaseScore": 0, "nutritionScore": 0,
+        "gutAge": "", "gutType": "" }},
+    "disease_risks": [{{"name": "", "value": "", "status": ""}}],
+    "nutrition": [{{"name": "", "value": "", "status": ""}}],
+    "amino_acids": [{{"name": "", "value": "", "status": ""}}],
+    "vitamins": [{{"name": "", "value": "", "status": ""}}],
+    "trace_elements": [{{"name": "", "value": "", "status": ""}}]
+}}
+
+只返回JSON,不要其他文字。"""
+
+            if self.llm_api_key:
+                # 调 OpenAI 兼容 API
+                import httpx
+                async with httpx.AsyncClient(timeout=60) as client:
+                    resp = await client.post(
+                        f"{settings.llm_base_url}/chat/completions",
+                        json={
+                            "model": settings.llm_model or "gpt-4o",
+                            "messages": [{"role": "user", "content": prompt}],
+                            "temperature": 0.1,
+                        },
+                        headers={"Authorization": f"Bearer {self.llm_api_key}"},
+                    )
+                    resp.raise_for_status()
+                    data = resp.json()
+                    content = data['choices'][0]['message']['content']
+                    content = content.replace('```json', '').replace('```', '').strip()
+                    return json.loads(content)
+            else:
+                logger.warning("LLM 未配置,跳过 LLM 兜底")
+                return None
+        except Exception as e:
+            logger.warning("LLM 解析失败: %s", e)
+            return None

+ 320 - 0
cfc-langgraph/app/api/adapter.py

@@ -0,0 +1,320 @@
+from fastapi import APIRouter, HTTPException
+from pydantic import BaseModel, ConfigDict
+from typing import Optional
+from app.graphs.chat_graph import create_chat_graph
+from app.graphs.analysis_graph import create_analysis_graph
+from app.graphs.health_coach_graph import create_health_coach_graph
+from app.graphs.health_butler_graph import create_health_butler_graph
+import uuid
+import time
+
+router = APIRouter(prefix="/api/v1", tags=["adapter"])
+
+# ---- Dify-compatible request/response models ----
+
+class DifyMessage(BaseModel):
+    role: str
+    content: str
+
+class DifyChatRequest(BaseModel):
+    # 前端 userId/conversationId 可能为数字,允许 int/float 自动转 str,避免 422
+    model_config = ConfigDict(coerce_numbers_to_str=True)
+    query: str = ""
+    user: str = "anonymous"
+    conversation_id: str = ""
+    messages: list[DifyMessage] = []
+    inputs: dict = {}
+    response_mode: str = "blocking"  # blocking / streaming
+    user_id: str = ""
+    bot_name: str = "AI管家"
+
+class DifyAnalysisRequest(BaseModel):
+    # 前端 userId 可能为数字,允许 int/float 自动转 str,避免 422
+    model_config = ConfigDict(coerce_numbers_to_str=True)
+    report_id: Optional[int] = None
+    user_id: Optional[str] = None
+    focus: Optional[str] = None
+    messages: list[DifyMessage] = []
+    inputs: dict = {}
+
+class HealthCoachGenerateRequest(BaseModel):
+    """Java 后端健康方案生成专用请求体"""
+    family_id: Optional[int] = None
+    member_ids: Optional[str] = None
+    dimensions: Optional[str] = None
+    goal: str = ""
+
+class DifyChoice(BaseModel):
+    index: int
+    message: dict
+    finish_reason: str = "stop"
+
+class DifyUsage(BaseModel):
+    prompt_tokens: int = 0
+    completion_tokens: int = 0
+    total_tokens: int = 0
+
+class DifyResponse(BaseModel):
+    id: str
+    object: str = "chat.completion"
+    created: int
+    model: str = "langgraph-cfc"
+    choices: list[DifyChoice]
+    usage: DifyUsage = DifyUsage()
+    metadata: dict = {}
+
+# ---- Helpers ----
+
+def _now_ts() -> int:
+    return int(time.time())
+
+def _extract_query(req: DifyChatRequest) -> str:
+    if req.query:
+        return req.query
+    # fallback: take last user message
+    for msg in reversed(req.messages):
+        if msg.role == "user":
+            return msg.content
+    return ""
+
+def _to_langgraph_context(req: DifyChatRequest) -> dict:
+    ctx = {}
+    if isinstance(req.inputs, dict):
+        ctx["child_id"] = req.inputs.get("child_id")
+        ctx["report_id"] = req.inputs.get("report_id")
+        ctx["family_id"] = req.inputs.get("family_id")
+    return ctx
+
+# ---- Dify-compatible endpoints ----
+
+@router.post("/chat/completion", response_model=DifyResponse)
+async def chat_completion(req: DifyChatRequest):
+    query = _extract_query(req)
+    if not query:
+        raise HTTPException(status_code=400, detail="query 为空")
+
+    graph = create_chat_graph()
+    initial_state = {
+        "query": query,
+        "user_id": int(req.user_id) if str(req.user_id).isdigit() else 0,
+        "conversation_id": req.conversation_id or None,
+        "intent": None,
+        "context": _to_langgraph_context(req),
+        "messages": None,
+        "answer": None,
+        "tasks": [],
+        "sources": [],
+    }
+    config = {
+        "configurable": {"thread_id": req.conversation_id or str(req.user_id or "default")},
+    }
+    result = await graph.ainvoke(initial_state, config)
+    answer = result.get("answer") or ""
+    return DifyResponse(
+        id=f"chatcmpl-{uuid.uuid4().hex[:24]}",
+        created=_now_ts(),
+        choices=[
+            DifyChoice(
+                index=0,
+                message={"role": "assistant", "content": answer},
+                finish_reason="stop",
+            )
+        ],
+        usage=DifyUsage(
+            prompt_tokens=len(query.split()),
+            completion_tokens=len(answer.split()),
+            total_tokens=len(query.split()) + len(answer.split()),
+        ),
+    )
+
+
+@router.post("/analysis/run", response_model=DifyResponse)
+async def analysis_run(req: DifyAnalysisRequest):
+    graph = create_analysis_graph()
+    report_id = req.report_id
+    if report_id is None and isinstance(req.inputs, dict):
+        report_id = req.inputs.get("report_id")
+    initial_state = {
+        "report_id": report_id,
+        "user_id": int(req.user_id) if str(req.user_id).isdigit() else 0,
+        "focus": req.focus,
+        "report_data": None,
+        "survey_data": None,
+        "dimension_scores": None,
+        "analysis": None,
+        "recommendations": [],
+    }
+    result = await graph.ainvoke(initial_state, {})
+    analysis = result.get("analysis") or ""
+    recommendations = result.get("recommendations") or []
+    content = analysis
+    if recommendations:
+        content += "\n\n建议:\n" + "\n".join(f"- {r}" for r in recommendations)
+    return DifyResponse(
+        id=f"analysis-{uuid.uuid4().hex[:24]}",
+        created=_now_ts(),
+        choices=[
+            DifyChoice(
+                index=0,
+                message={"role": "assistant", "content": content},
+                finish_reason="stop",
+            )
+        ],
+        usage=DifyUsage(
+            prompt_tokens=len((req.focus or "").split()),
+            completion_tokens=len(content.split()),
+            total_tokens=len((req.focus or "").split()) + len(content.split()),
+        ),
+    )
+
+
+@router.post("/health/coach", response_model=DifyResponse)
+async def health_coach(req: DifyChatRequest):
+    query = _extract_query(req)
+    if not query:
+        raise HTTPException(status_code=400, detail="query 为空")
+
+    graph = create_health_coach_graph()
+    initial_state = {
+        "query": query,
+        "user_id": int(req.user_id) if str(req.user_id).isdigit() else 0,
+        "child_id": int(req.inputs.get("child_id")) if req.inputs.get("child_id") else None,
+        "conversation_id": req.conversation_id or None,
+        "context": _to_langgraph_context(req),
+        "answer": None,
+        "sources": [],
+        "memory_messages": None,
+    }
+    config = {
+        "configurable": {"thread_id": req.conversation_id or f"health_{req.user_id}"},
+    }
+    result = await graph.ainvoke(initial_state, config)
+    answer = result.get("answer") or ""
+
+    sources = result.get("sources") or []
+    metadata = {}
+    if sources:
+        metadata["sources"] = [
+            {"title": s.get("title", ""), "type": s.get("type", "knowledge")}
+            for s in sources
+        ]
+
+    return DifyResponse(
+        id=f"health-{uuid.uuid4().hex[:24]}",
+        created=_now_ts(),
+        model="langgraph-health-coach",
+        choices=[
+            DifyChoice(
+                index=0,
+                message={"role": "assistant", "content": answer},
+                finish_reason="stop",
+            )
+        ],
+        usage=DifyUsage(
+            prompt_tokens=len(query.split()),
+            completion_tokens=len(answer.split()),
+            total_tokens=len(query.split()) + len(answer.split()),
+        ),
+        metadata=metadata,
+    )
+
+
+@router.post("/health/butler", response_model=DifyResponse)
+async def health_butler(req: DifyChatRequest):
+    """AI 健康管家 — 基于健康知识库检索 + 个性化建议 + 任务生成"""
+    query = _extract_query(req)
+    if not query:
+        raise HTTPException(status_code=400, detail="query 为空")
+
+    graph = create_health_butler_graph()
+
+    # 从 inputs 中提取健康管家需要的上下文
+    inputs = req.inputs or {}
+    family_id = inputs.get("family_id")
+    child_id = inputs.get("child_id")
+    report_id = inputs.get("report_id")
+    focus = inputs.get("focus")
+
+    # conversation_id 用于 checkpointer thread
+    thread_id = req.conversation_id or f"butler_{req.user_id or 'anon'}"
+
+    initial_state = {
+        "query": query,
+        "user_id": int(req.user_id) if str(req.user_id).isdigit() else 0,
+        "conversation_id": thread_id,
+        "family_id": int(family_id) if family_id else None,
+        "child_id": int(child_id) if child_id else None,
+        "report_id": int(report_id) if report_id else None,
+        "focus": focus,
+        "kb_context": None,
+        "knowledge_results": [],
+        "answer": None,
+        "tasks": [],
+        "sources": [],
+        "messages": None,
+    }
+    config = {
+        "configurable": {"thread_id": thread_id},
+    }
+    result = await graph.ainvoke(initial_state, config)
+    answer = result.get("answer") or ""
+
+    # 从回答中提取 TASK 标记
+    tasks = result.get("tasks") or []
+    sources = result.get("sources") or []
+    metadata = {"tasks": tasks, "sources": sources}
+
+    return DifyResponse(
+        id=f"butler-{uuid.uuid4().hex[:24]}",
+        created=_now_ts(),
+        model="langgraph-health-butler",
+        choices=[
+            DifyChoice(
+                index=0,
+                message={"role": "assistant", "content": answer},
+                finish_reason="stop",
+            )
+        ],
+        usage=DifyUsage(
+            prompt_tokens=len(query.split()),
+            completion_tokens=len(answer.split()),
+            total_tokens=len(query.split()) + len(answer.split()),
+        ),
+        metadata=metadata,
+    )
+
+
+@router.post("/health/coach/generate")
+async def health_coach_generate(req: HealthCoachGenerateRequest):
+    """Java 后端健康方案生成端点 — 接收 familyId/memberIds/dimensions/goal,返回方案文本"""
+    graph = create_health_coach_graph()
+    goal = req.goal or "生成健康方案"
+    family_id = req.family_id
+    member_ids = req.member_ids
+    dimensions = req.dimensions
+
+    # 构造用户查询:将目标和维度信息注入 prompt
+    query_parts = [goal]
+    if dimensions:
+        query_parts.append(f"重点关注维度:{dimensions}")
+    if member_ids:
+        query_parts.append(f"目标成员IDs:{member_ids}")
+    query = ";".join(query_parts)
+
+    initial_state = {
+        "query": query,
+        "user_id": family_id or 0,
+        "child_id": None,
+        "conversation_id": None,
+        "context": {
+            "family_id": family_id,
+            "member_ids": member_ids,
+            "dimensions": dimensions,
+        },
+        "answer": None,
+        "sources": [],
+        "memory_messages": None,
+    }
+    result = await graph.ainvoke(initial_state, {})
+    answer = result.get("answer") or ""
+    return answer

+ 45 - 0
cfc-langgraph/app/api/analyze.py

@@ -0,0 +1,45 @@
+from fastapi import APIRouter
+from pydantic import BaseModel
+from typing import Optional
+from app.graphs.analysis_graph import create_analysis_graph
+
+router = APIRouter(prefix="/api/v1", tags=["analyze"])
+
+_graph = None
+
+
+def get_graph():
+    global _graph
+    if _graph is None:
+        _graph = create_analysis_graph()
+    return _graph
+
+
+class AnalyzeRequest(BaseModel):
+    report_id: int
+    user_id: int
+    focus: Optional[str] = None
+
+
+class AnalyzeResponse(BaseModel):
+    analysis: str = ""
+    recommendations: list[str] = []
+    trace_id: str = ""
+
+
+@router.post("/analyze", response_model=AnalyzeResponse)
+async def analyze(req: AnalyzeRequest):
+    graph = get_graph()
+    import uuid
+    trace_id = str(uuid.uuid4())
+    result = await graph.ainvoke({
+        "report_id": req.report_id,
+        "user_id": req.user_id,
+        "focus": req.focus,
+        "report_data": None,
+        "survey_data": None,
+        "dimension_scores": None,
+        "analysis": None,
+        "recommendations": [],
+    })
+    return AnalyzeResponse(analysis=result.get("analysis", ""), trace_id=trace_id)

+ 58 - 0
cfc-langgraph/app/api/chat.py

@@ -0,0 +1,58 @@
+import uuid
+from fastapi import APIRouter
+from app.models.chat import ChatRequest, ChatResponse, SourceInfo
+from app.graphs.chat_graph import create_chat_graph
+
+router = APIRouter(prefix="/api/v1", tags=["chat"])
+
+_graph = None
+
+
+def get_graph():
+    global _graph
+    if _graph is None:
+        _graph = create_chat_graph()
+    return _graph
+
+
+@router.post("/chat", response_model=ChatResponse)
+async def chat(req: ChatRequest):
+    """家庭聊天: 意图分类→上下文→LLM→记忆"""
+    trace_id = str(uuid.uuid4())
+    graph = get_graph()
+
+    initial_state = {
+        "query": req.query,
+        "user_id": req.user_id,
+        "conversation_id": req.conversation_id,
+        "child_id": req.context.child_id if req.context else None,
+        "intent": None,
+        "context": None,
+        "messages": None,
+        "answer": None,
+        "tasks": [],
+        "sources": [],
+    }
+
+    config = {
+        "configurable": {"thread_id": req.conversation_id or str(req.user_id)},
+    }
+
+    result = await graph.ainvoke(initial_state, config)
+
+    sources = []
+    for s in result.get("sources", []):
+        sources.append(SourceInfo(
+            type=s.get("type", "tool"),
+            title=s.get("name", ""),
+        ))
+
+    conv_id = req.conversation_id or f"conv_{req.user_id}_{__import__('time').time()}"
+
+    return ChatResponse(
+        answer=result.get("answer", ""),
+        conversation_id=conv_id,
+        sources=sources,
+        tasks=result.get("tasks", []),
+        trace_id=trace_id,
+    )

+ 38 - 0
cfc-langgraph/app/api/logs.py

@@ -0,0 +1,38 @@
+"""请求日志查询接口"""
+from fastapi import APIRouter
+from app.middleware import get_request_log, clear_request_log
+
+router = APIRouter(tags=["logs"])
+
+
+@router.get("/api/v1/logs")
+async def list_logs(
+    limit: int = 100,
+    path: str = None,
+    method: str = None,
+    status: int = None,
+):
+    """
+    获取请求日志(最近 N 条)
+    - limit: 返回条数(默认100,最大500)
+    - path: 路径过滤(支持模糊)
+    - method: 方法过滤(GET/POST)
+    - status: 状态码过滤
+    """
+    logs = get_request_log()
+    if limit and 0 < limit <= 500:
+        logs = logs[:limit]
+    if path:
+        logs = [l for l in logs if path.lower() in l.get("path", "").lower()]
+    if method:
+        logs = [l for l in logs if l.get("method", "").upper() == method.upper()]
+    if status:
+        logs = [l for l in logs if l.get("status") == status]
+    return {"total": len(get_request_log()), "count": len(logs), "logs": logs}
+
+
+@router.post("/api/v1/logs/clear")
+async def clear_logs():
+    """清空请求日志"""
+    n = clear_request_log()
+    return {"cleared": n}

+ 135 - 0
cfc-langgraph/app/api/meal.py

@@ -0,0 +1,135 @@
+import uuid
+import json
+import logging
+import re
+from fastapi import APIRouter
+from app.models.meal import (
+    FoodRecognizeRequest,
+    FoodRecognizeResponse,
+    FoodsItem,
+    MenuGenerateRequest,
+    MenuGenerateResponse,
+)
+from app.config import settings
+
+logger = logging.getLogger(__name__)
+router = APIRouter(prefix="/api/v1", tags=["meal"])
+
+
+@router.post("/food/recognize", response_model=FoodRecognizeResponse)
+async def recognize_food(req: FoodRecognizeRequest):
+    """食材识别: 上传图片 URL, 返回识别食材列表"""
+    trace_id = str(uuid.uuid4())
+    try:
+        from langchain_openai import ChatOpenAI
+        from langchain_core.messages import HumanMessage, SystemMessage
+
+        llm = ChatOpenAI(
+            model=settings.llm_model,
+            api_key=settings.llm_api_key,
+            base_url=settings.llm_base_url,
+            temperature=0,
+        )
+
+        prompt = f"""请识别以下图片中的主要食材(3-8种),按置信度排序。
+返回 JSON 数组格式,不要包含其他内容:
+[{{"name": "食材名称", "confidence": 0.95, "category": "蔬菜/水果/肉禽/水产/蛋奶/谷物/调味/其他"}}]
+
+图片链接: {req.image_url}
+要求: 只返回 JSON 数组,不要有任何其他文字。"""
+
+        response = llm.invoke([
+            SystemMessage(content="你是一个专业的食材识别助手。"),
+            HumanMessage(content=prompt),
+        ])
+
+        text = response.content.strip()
+        try:
+            foods = json.loads(text)
+        except json.JSONDecodeError:
+            match = re.search(r'\[[\s\S]*\]', text)
+            foods = json.loads(match.group()) if match else []
+
+        foods_list = []
+        for item in foods:
+            if isinstance(item, dict):
+                foods_list.append(FoodsItem(
+                    name=item.get("name", "未知"),
+                    confidence=float(item.get("confidence", 0.5)),
+                    category=item.get("category", "other"),
+                ))
+
+        return FoodRecognizeResponse(foods=foods_list, raw_response=text, trace_id=trace_id)
+    except Exception as e:
+        logger.error("食材识别失败: %s", e, exc_info=True)
+        return FoodRecognizeResponse(foods=[], trace_id=trace_id)
+
+
+@router.post("/menu/generate", response_model=MenuGenerateResponse)
+async def generate_menu(req: MenuGenerateRequest):
+    """菜单生成: 根据食材和用餐人数生成一日三餐菜单"""
+    trace_id = str(uuid.uuid4())
+    try:
+        from langchain_openai import ChatOpenAI
+        from langchain_core.messages import HumanMessage, SystemMessage
+
+        try:
+            selected_foods = json.loads(req.selected_foods) if req.selected_foods else []
+        except json.JSONDecodeError:
+            selected_foods = []
+
+        try:
+            participants = json.loads(req.participants) if req.participants else []
+        except json.JSONDecodeError:
+            participants = []
+
+        participant_count = len(participants) if participants else 1
+        foods_text = ", ".join([f.get("name", "") for f in selected_foods]) if selected_foods else "根据可用食材"
+
+        constraints = []
+        if req.allergies:
+            constraints.append(f"禁忌: {req.allergies}")
+        if req.health_goals:
+            constraints.append(f"健康目标: {req.health_goals}")
+        if req.cuisine_pref:
+            constraints.append(f"菜系偏好: {req.cuisine_pref}")
+        if req.spice_level is not None:
+            constraints.append(f"辣度: {req.spice_level}/5")
+        constraints_text = "\n".join(constraints) if constraints else "无特殊限制"
+
+        prompt = f"""请为{participant_count}人生成{req.date}的一日三餐菜单。
+
+可用食材: {foods_text}
+用餐人数: {participant_count}人
+{constraints_text}
+
+请返回 JSON 格式:
+{{"meals": [{{"type": "breakfast", "name": "早餐", "dishes": [{{"name": "菜品名", "ingredients": [{{"name": "食材", "grams": 100}}], "cooking_method": "烹饪方法", "nutrition": {{"calories": 200, "protein": 10, "carbs": 30, "fat": 5}}, "notes": "备注"}}]}}]}}
+
+要求:
+1. 早/午/晚各至少1-2道菜
+2. 食材用量按{participant_count}人份计算
+3. 营养均衡,考虑健康目标
+4. 只用提供的食材
+5. 只返回 JSON,不要其他文字"""
+
+        llm = ChatOpenAI(
+            model=settings.llm_model,
+            api_key=settings.llm_api_key,
+            base_url=settings.llm_base_url,
+            temperature=0.7,
+        )
+
+        response = llm.invoke([
+            SystemMessage(content="你是一个专业营养师和厨师,擅长根据食材和健康目标设计食谱。"),
+            HumanMessage(content=prompt),
+        ])
+
+        text = response.content.strip()
+        match = re.search(r'\{[\s\S]*\}', text)
+        menu_json = match.group() if match else '{"meals": []}'
+
+        return MenuGenerateResponse(menu_json=menu_json, trace_id=trace_id)
+    except Exception as e:
+        logger.error("菜单生成失败: %s", e, exc_info=True)
+        return MenuGenerateResponse(menu_json='{"meals": []}', trace_id=trace_id)

+ 42 - 0
cfc-langgraph/app/api/recommend.py

@@ -0,0 +1,42 @@
+import uuid
+from fastapi import APIRouter
+from app.models.recommend import RecommendRequest, RecommendResponse, RecommendItem
+from app.graphs.recommend_graph import RecommendAgent
+import logging
+
+logger = logging.getLogger(__name__)
+router = APIRouter(prefix="/api/v1", tags=["recommend"])
+
+_agent: RecommendAgent = None
+
+
+def get_agent() -> RecommendAgent:
+    global _agent
+    if _agent is None:
+        _agent = RecommendAgent()
+    return _agent
+
+
+@router.post("/recommend", response_model=RecommendResponse)
+async def recommend(req: RecommendRequest):
+    """营养推荐: Agent 搜索+LLM 解释"""
+    trace_id = str(uuid.uuid4())
+    try:
+        agent = get_agent()
+        result = await agent.run(query=req.query, tags=req.tags, limit=req.limit)
+
+        items_data = result.get("items", [])
+        items = []
+        for item in items_data:
+            items.append(RecommendItem(
+                type=item.get("type", "product"),
+                id=item.get("id", 0),
+                name=item.get("name", ""),
+                description=item.get("description", ""),
+                reason=item.get("reason", ""),
+            ))
+
+        return RecommendResponse(items=items, source="agent", trace_id=trace_id)
+    except Exception as e:
+        logger.error("RecommendAgent 调用失败: %s", e, exc_info=True)
+        return RecommendResponse(items=[], source="error", trace_id=trace_id)

+ 4 - 23
cfc-langgraph/app/api/report_parse.py

@@ -5,7 +5,6 @@ from app.agents.report_parse_agent import ReportParseAgent
 import logging
 import os
 import json
-import tempfile
 
 logger = logging.getLogger(__name__)
 router = APIRouter(prefix="/api/v1", tags=["report_parse"])
@@ -31,34 +30,16 @@ class ParseResponse(BaseModel):
     data: dict = {}
 
 
-async def resolve_local_path(file_path: str) -> str:
-    """将 file_path 解析为本地文件路径。
-    支持:本地路径、/uploads 共享卷路径、http(s) 远程 OSS URL(下载到临时文件)。"""
-    if file_path.startswith(('http://', 'https://')):
-        import httpx
-        logger.info("report_parse: 远程文件 %s,下载到临时文件", file_path)
-        async with httpx.AsyncClient(timeout=120, follow_redirects=True) as client:
-            resp = await client.get(file_path)
-            resp.raise_for_status()
-        ext = os.path.splitext(file_path.split('?')[0])[1] or '.pdf'
-        fd, tmp = tempfile.mkstemp(suffix=ext)
-        with os.fdopen(fd, 'wb') as f:
-            f.write(resp.content)
-        logger.info("report_parse: 下载完成 -> %s", tmp)
-        return tmp
-    return file_path
-
-
 @router.post("/report/parse", response_model=ParseResponse)
 async def parse_report(req: ParseRequest):
-    local = await resolve_local_path(req.file_path)
-    if not os.path.exists(local):
+    if not os.path.exists(req.file_path):
         raise HTTPException(status_code=400, detail=f"文件不存在: {req.file_path}")
 
-    logger.info("report_parse: file_path=%s family_id=%s user_id=%s", local, req.family_id, req.user_id)
+    logger.info("report_parse: file_path=%s family_id=%s user_id=%s",
+                req.file_path, req.family_id, req.user_id)
     agent = get_agent()
     try:
-        result = await agent.parse(local)
+        result = await agent.parse(req.file_path)
         return ParseResponse(data=result)
     except Exception as e:
         logger.error("报告解析失败: %s", e, exc_info=True)

+ 45 - 0
cfc-langgraph/app/api/tongue.py

@@ -0,0 +1,45 @@
+from fastapi import APIRouter, UploadFile, File, Form
+from typing import Optional
+from app.agents.multimodal_agent import TongueDiagnosisAgent
+import logging
+
+logger = logging.getLogger(__name__)
+router = APIRouter(prefix="/api/v1", tags=["tongue"])
+
+_agent: Optional[TongueDiagnosisAgent] = None
+
+
+def get_agent() -> TongueDiagnosisAgent:
+    global _agent
+    if _agent is None:
+        _agent = TongueDiagnosisAgent()
+    return _agent
+
+
+@router.post("/tongue/diagnose")
+async def tongue_diagnose(
+    file: UploadFile = File(...),
+    user_id: int = Form(...),
+):
+    """舌诊分析: 上传舌苔图片, 返回分析结果"""
+    agent = get_agent()
+
+    import tempfile, os
+    ext = os.path.splitext(file.filename or "tongue.jpg")[1] or ".jpg"
+    tmp = tempfile.NamedTemporaryFile(delete=False, suffix=ext)
+    content = await file.read()
+    tmp.write(content)
+    tmp.close()
+
+    try:
+        import base64
+        b64 = base64.b64encode(content).decode()
+        data_url = f"data:image/{ext[1:]};base64,{b64}"
+
+        result = await agent.diagnose(image_url=data_url, user_id=user_id)
+        return {"code": 200, "data": result}
+    except Exception as e:
+        logger.error("舌诊分析失败: %s", e, exc_info=True)
+        return {"code": 500, "message": "舌诊分析失败"}
+    finally:
+        os.unlink(tmp.name)

+ 96 - 0
cfc-langgraph/app/graphs/analysis_graph.py

@@ -0,0 +1,96 @@
+from typing import TypedDict, Optional
+from langgraph.graph import StateGraph, START, END
+from langgraph.checkpoint.memory import MemorySaver
+from langchain_openai import ChatOpenAI
+from langchain_core.messages import SystemMessage, HumanMessage
+from app.tools.report_tools import get_report_detail, get_survey_data, get_dimension_scores
+from app.config import settings
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class AnalysisState(TypedDict):
+    report_id: int
+    user_id: int
+    focus: Optional[str]          # nutrition / gut / chronic / overall
+    report_data: Optional[dict]
+    survey_data: Optional[dict]
+    dimension_scores: Optional[dict]
+    analysis: Optional[str]
+    recommendations: list[str]
+
+
+ANALYSIS_SYSTEM_PROMPT = """你是一个儿童健康报告解读专家。根据健康报告数据和调研问卷, 提供专业、易懂的分析。
+
+分析原则:
+1. 用通俗语言解释各项指标含义
+2. 关注异常指标, 给出改善建议
+3. 结合问卷数据提供个性化分析
+4. 五维能量(身/心/智/行/富)角度解读整体状况
+5. 输出格式: 总体评估 → 分项分析 → 改善建议
+"""
+
+
+def create_analysis_graph():
+    """创建报告解读 StateGraph"""
+    llm = ChatOpenAI(
+        model=settings.llm_model,
+        api_key=settings.llm_api_key,
+        base_url=settings.llm_base_url,
+        temperature=settings.llm_temperature,
+    )
+    llm_with_tools = llm.bind_tools([
+        get_report_detail,
+        get_survey_data,
+        get_dimension_scores,
+    ])
+
+    builder = StateGraph(AnalysisState)
+
+    async def gather_data(state: AnalysisState) -> dict:
+        """收集报告 + 问卷 + 维度数据"""
+        result = {}
+        if state.get("report_id"):
+            try:
+                detail = await get_report_detail.ainvoke({"report_id": state["report_id"]})
+                result["report_data"] = detail
+            except Exception as e:
+                logger.warning("获取报告详情失败: %s", e)
+            try:
+                survey = await get_survey_data.ainvoke({"report_id": state["report_id"]})
+                result["survey_data"] = survey
+            except Exception as e:
+                logger.warning("获取问卷数据失败: %s", e)
+        return result
+
+    async def analyze(state: AnalysisState) -> dict:
+        """LLM 分析"""
+        context_parts = []
+        if state.get("report_data"):
+            context_parts.append(f"报告数据: {state['report_data']}")
+        if state.get("survey_data"):
+            context_parts.append(f"问卷数据: {state['survey_data']}")
+        if state.get("focus"):
+            context_parts.append(f"重点关注: {state['focus']}")
+
+        context_text = "\n".join(context_parts) if context_parts else "暂无数据"
+
+        messages = [
+            SystemMessage(content=ANALYSIS_SYSTEM_PROMPT),
+            SystemMessage(content=f"待分析数据:\n{context_text}"),
+            HumanMessage(content="请分析以上健康数据, 给出评估和建议。"),
+        ]
+
+        response = await llm_with_tools.ainvoke(messages)
+        return {"analysis": response.content}
+
+    builder.add_node("gather_data", gather_data)
+    builder.add_node("analyze", analyze)
+
+    builder.add_edge(START, "gather_data")
+    builder.add_edge("gather_data", "analyze")
+    builder.add_edge("analyze", END)
+
+    checkpointer = MemorySaver()
+    return builder.compile(checkpointer=checkpointer)

+ 178 - 0
cfc-langgraph/app/graphs/chat_graph.py

@@ -0,0 +1,178 @@
+from typing import TypedDict, Literal
+from langgraph.graph import StateGraph, START, END
+from langgraph.checkpoint.memory import MemorySaver
+from langchain_openai import ChatOpenAI
+from langchain_core.messages import SystemMessage, HumanMessage
+from app.agents.intent_classifier import IntentClassifier, Intent
+from app.agents.chat_agent import ChatAgent
+from app.tools.product_tools import (
+    search_product_by_keyword,
+    search_article_by_keyword,
+    search_activity_by_keyword,
+)
+from app.memory.store import MemoryManager
+from app.config import settings
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class ChatState(TypedDict):
+    query: str
+    user_id: int
+    conversation_id: str
+    child_id: int | None
+    intent: Intent | None
+    context: dict | None
+    messages: list | None
+    answer: str | None
+    tasks: list[dict]
+    sources: list[dict]
+
+
+CHAT_SYSTEM_PROMPT = """你是一个儿童成长家庭助手, 回答关于孩子成长、健康、教育的各种问题。
+
+你可以使用搜索工具查找商品、活动和文章来辅助回答。
+
+回答原则:
+1. 用中文, 语气温暖亲切
+2. 如果用户提到具体孩子, 参考提供的家庭上下文
+3. 需要推荐时使用搜索工具
+4. 可以生成 [TASK: {"title": "任务名", "dimension": "身/心/智/行/富", "points": 10}] 标记来创建行动任务
+5. 不要编造医疗建议, 严重问题建议咨询医生
+"""
+
+
+def create_chat_graph():
+    """创建聊天 StateGraph"""
+    agent = ChatAgent()
+    classifier = IntentClassifier()
+
+    llm = ChatOpenAI(
+        model=settings.llm_model,
+        api_key=settings.llm_api_key,
+        base_url=settings.llm_base_url,
+        temperature=settings.llm_temperature,
+    )
+    
+    # 初始化嵌入模型用于记忆存储
+    from langchain_openai import OpenAIEmbeddings
+    embeddings = OpenAIEmbeddings(
+        model=settings.embedding_model,
+        api_key=settings.effective_embedding_api_key,
+        base_url=settings.effective_embedding_base_url,
+    )
+    memory_mgr = MemoryManager(embeddings)
+
+    llm_with_tools = llm.bind_tools([
+        search_product_by_keyword,
+        search_article_by_keyword,
+        search_activity_by_keyword,
+    ])
+
+    builder = StateGraph(ChatState)
+
+    # ── 节点 ──
+
+    async def classify_intent(state: ChatState) -> dict:
+        intent = await classifier.classify(
+            state["query"],
+            context=str(state.get("context", {})),
+        )
+        return {"intent": intent}
+
+    async def load_context(state: ChatState) -> dict:
+        ctx = await agent.load_context(state["user_id"], state.get("child_id"))
+        return {"context": ctx}
+
+    async def llm_call(state: ChatState) -> dict:
+        """核心 LLM 调用 + Tool"""
+        messages = [SystemMessage(content=CHAT_SYSTEM_PROMPT)]
+
+        # 注入家庭上下文
+        ctx = state.get("context", {})
+        if ctx:
+            ctx_text = f"\n家庭上下文:\n{ctx}"
+            messages.append(SystemMessage(content=ctx_text))
+
+        # 注入长期记忆
+        try:
+            memories = await memory_mgr.recall(state["user_id"], state["query"])
+            if memories:
+                mem_text = "\n".join([f"- {m}" for m in memories])
+                messages.append(SystemMessage(
+                    content=f"相关历史对话:\n{mem_text}"
+                ))
+        except Exception as e:
+            logger.warning("召回记忆失败: %s", e)
+
+        # 用户消息
+        messages.append(HumanMessage(content=state["query"]))
+
+        response = await llm_with_tools.ainvoke(messages)
+        answer = response.content
+
+        # 提取任务
+        tasks = await agent.extract_tasks(answer, state["user_id"], state["conversation_id"])
+
+        # 提取来源
+        sources = []
+        if response.response_metadata.get("tool_calls"):
+            for tc in response.response_metadata["tool_calls"]:
+                sources.append({
+                    "type": "tool",
+                    "name": tc.get("name", ""),
+                    "input": tc.get("args", {}),
+                })
+
+        return {
+            "answer": answer,
+            "tasks": tasks,
+            "sources": sources,
+            "messages": [{"role": "user", "content": state["query"]},
+                         {"role": "assistant", "content": answer}],
+        }
+
+    async def save_memory(state: ChatState) -> dict:
+        """对话后保存到长期记忆"""
+        try:
+            if state.get("messages"):
+                await memory_mgr.save_conversation(
+                    state["user_id"],
+                    state["conversation_id"],
+                    state["messages"],
+                )
+        except Exception as e:
+            logger.warning("保存记忆失败: %s", e)
+        return {}
+
+    # ── 路由 ──
+
+    def route_by_intent(state: ChatState) -> Literal["llm_call", END]:
+        if state["intent"] in (
+            Intent.RECOMMEND,
+            Intent.ANALYSIS,
+            Intent.HEALTH,
+        ):
+            # 这些意图需要更专业的 Agent (Phase 3 实现)
+            # 当前先走通用 LLM
+            pass
+        return "llm_call"
+
+    # ── 构建图 ──
+
+    builder.add_node("classify_intent", classify_intent)
+    builder.add_node("load_context", load_context)
+    builder.add_node("llm_call", llm_call)
+    builder.add_node("save_memory", save_memory)
+
+    builder.add_edge(START, "classify_intent")
+    builder.add_edge("classify_intent", "load_context")
+    builder.add_conditional_edges("load_context", route_by_intent)
+    builder.add_edge("llm_call", "save_memory")
+    builder.add_edge("save_memory", END)
+
+    checkpointer = MemorySaver()
+    graph = builder.compile(checkpointer=checkpointer)
+
+    return graph

+ 322 - 0
cfc-langgraph/app/graphs/health_butler_graph.py

@@ -0,0 +1,322 @@
+from typing import TypedDict, Optional, Literal
+from langgraph.graph import StateGraph, START, END
+from langgraph.checkpoint.memory import MemorySaver
+from langchain_openai import ChatOpenAI, OpenAIEmbeddings
+from langchain_core.messages import SystemMessage, HumanMessage
+from app.tools.java_client import JavaClient
+from app.tools.product_tools import (
+    search_product_by_keyword,
+    search_article_by_keyword,
+)
+from app.memory.store import MemoryManager
+from app.config import settings
+import logging
+import json
+
+logger = logging.getLogger(__name__)
+
+
+class HealthButlerState(TypedDict):
+    query: str
+    user_id: int
+    conversation_id: str
+    family_id: int | None
+    child_id: int | None
+    report_id: int | None
+    focus: str | None          # gut / nutrition / overall / chronic
+    kb_context: dict | None
+    knowledge_results: list[dict]
+    answer: str | None
+    tasks: list[dict]
+    sources: list[dict]
+    messages: list | None
+
+
+HEALTH_BUTLER_PROMPT = """你是一个家庭AI健康管家, 为用户提供专业、温暖的个性化健康建议。
+
+你的能力:
+1. 基于健康知识库(菌属功能、指标解读、营养素、食物)提供科学建议
+2. 解读健康报告数据, 结合知识库给出改善方案
+3. 根据菌群检测结果推荐精准营养方案
+4. 提供可执行的健康微行动
+
+回答原则:
+1. 用中文, 像家庭医生一样亲切但专业
+2. 引用知识库中的菌属/指标/食物信息作为证据
+3. 优先使用知识库数据, 其次使用通用健康知识
+4. 需要产品推荐时使用搜索工具
+5. 生成 [TASK: {"title": "任务名", "dimension": "身", "points": 10, "frequency": "每天"}] 标记创建健康行动任务
+6. 超出知识库范围的慢性病/急症问题, 建议咨询专业医生
+7. 不要编造医学建议, 不夸大效果
+"""
+
+
+def create_health_butler_graph():
+    """创建 AI 健康管家 StateGraph — 知识检索 + LLM 建议 + 记忆"""
+    llm = ChatOpenAI(
+        model=settings.llm_model,
+        api_key=settings.llm_api_key,
+        base_url=settings.llm_base_url,
+        temperature=settings.llm_temperature,
+    )
+
+    embeddings = OpenAIEmbeddings(
+        model=settings.embedding_model,
+        api_key=settings.effective_embedding_api_key,
+        base_url=settings.effective_embedding_base_url,
+    )
+    memory_mgr = MemoryManager(embeddings)
+
+    java = JavaClient()
+
+    llm_with_tools = llm.bind_tools([
+        search_product_by_keyword,
+        search_article_by_keyword,
+    ])
+
+    builder = StateGraph(HealthButlerState)
+
+    async def load_knowledge(state: HealthButlerState) -> dict:
+        """阶段1: 从 Java 后端加载健康知识库上下文"""
+        context = {}
+        try:
+            client = await java._get_client()
+            resp = await client.post(
+                "/api/health/knowledge/list",
+                json={},
+            )
+            data = resp.json()
+            if data.get("code") == 200:
+                items = data.get("data", [])
+                # 按 itemType 分组
+                by_type = {}
+                for item in items:
+                    t = item.get("itemType", "unknown")
+                    if t not in by_type:
+                        by_type[t] = []
+                    by_type[t].append(item)
+                context["kb_items"] = len(items)
+                context["kb_by_type"] = {k: len(v) for k, v in by_type.items()}
+                context["kb_loaded"] = True
+            else:
+                context["kb_loaded"] = False
+        except Exception as e:
+            logger.warning("加载健康知识库失败: %s", e)
+            context["kb_loaded"] = False
+
+        return {"kb_context": context}
+
+    async def search_knowledge(state: HealthButlerState) -> dict:
+        """阶段2: 根据 query 意图精准搜索知识库"""
+        query = state["query"]
+        results = []
+
+        # 提取 query 中可能的菌属名/指标名/食物名关键词
+        # 策略: 对 query 做简单的关键词匹配, 调用 Java 知识库逐个查询
+        keywords = _extract_health_keywords(query)
+        if keywords:
+            client = await java._get_client()
+            for kw in keywords[:5]:  # 最多查询5个关键词
+                # 优先查 bacteria 类型
+                for item_type in ["bacteria", "indicator", "food", "nutrient"]:
+                    try:
+                        resp = await client.post(
+                            "/api/health/knowledge/query",
+                            json={"itemType": item_type, "itemName": kw},
+                        )
+                        data = resp.json()
+                        if data.get("code") == 200 and data.get("data"):
+                            results.append({
+                                "type": item_type,
+                                "keyword": kw,
+                                "data": data["data"],
+                            })
+                            break  # 命中即跳出
+                    except Exception:
+                        continue
+
+        return {"knowledge_results": results}
+
+    async def llm_answer(state: HealthButlerState) -> dict:
+        """阶段3: LLM 生成个性化健康建议"""
+        messages = [SystemMessage(content=HEALTH_BUTLER_PROMPT)]
+
+        # 注入健康知识库上下文
+        kb_ctx = state.get("kb_context", {})
+        knowledge_results = state.get("knowledge_results", [])
+        if knowledge_results:
+            kb_text_parts = []
+            for r in knowledge_results:
+                d = r.get("data", {})
+                name = d.get("itemName", r.get("keyword", ""))
+                cat = d.get("category", "")
+                desc = d.get("description", "")
+                suggestion = d.get("suggestion", "")
+                normal_range = d.get("normalRange", "")
+
+                parts = [f"【{r['type']}】{name}"]
+                if cat:
+                    parts.append(f"(分类: {cat})")
+                if normal_range:
+                    parts.append(f"正常范围: {normal_range}")
+                if desc:
+                    parts.append(f"{desc}")
+                if suggestion:
+                    parts.append(f"建议: {suggestion}")
+                kb_text_parts.append(" ".join(parts))
+
+            kb_text = "\n---\n".join(kb_text_parts)
+            messages.append(SystemMessage(content=f"健康知识库参考:\n{kb_text}"))
+
+        # 注入家庭上下文
+        family_id = state.get("family_id")
+        if family_id:
+            try:
+                client = await java._get_client()
+                resp = await client.post(
+                    "/api/ai/context",
+                    json={
+                        "user_id": str(state["user_id"]),
+                        "params": {
+                            "familyId": family_id,
+                            **({"childId": state["child_id"]} if state.get("child_id") else {}),
+                        },
+                    },
+                )
+                data = resp.json()
+                if data.get("code") == 200:
+                    ctx = data.get("data", {})
+                    if ctx:
+                        messages.append(SystemMessage(content=f"家庭上下文:\n{json.dumps(ctx, ensure_ascii=False, indent=2)}"))
+            except Exception as e:
+                logger.debug("加载家庭上下文跳过: %s", e)
+
+        # 注入长期记忆
+        try:
+            memories = await memory_mgr.recall(state["user_id"], state["query"])
+            if memories:
+                mem_text = "\n".join([f"- {m}" for m in memories])
+                messages.append(SystemMessage(content=f"相关历史对话:\n{mem_text}"))
+        except Exception as e:
+            logger.warning("召回记忆失败: %s", e)
+
+        # 用户消息
+        focus = state.get("focus")
+        query = state["query"]
+        if focus:
+            query = f"[重点关注: {focus}] {query}"
+        messages.append(HumanMessage(content=query))
+
+        response = await llm_with_tools.ainvoke(messages)
+        answer = response.content
+
+        # 提取任务
+        import re
+        tasks = []
+        pattern = r'\[TASK:\s*\{([^}]+)\}\]'
+        for match in re.findall(pattern, answer):
+            task_info = {}
+            for kv in match.split(","):
+                if ":" in kv:
+                    k, v = kv.split(":", 1)
+                    task_info[k.strip()] = v.strip().strip('"').strip("'")
+            if task_info.get("title"):
+                tasks.append(task_info)
+
+        sources = []
+        for r in knowledge_results:
+            d = r.get("data", {})
+            sources.append({
+                "type": "health_kb",
+                "name": d.get("itemName", r.get("keyword", "")),
+                "category": d.get("category", ""),
+            })
+        if response.response_metadata.get("tool_calls"):
+            for tc in response.response_metadata["tool_calls"]:
+                sources.append({
+                    "type": "tool",
+                    "name": tc.get("name", ""),
+                })
+
+        return {
+            "answer": answer,
+            "tasks": tasks,
+            "sources": sources,
+            "messages": [
+                {"role": "user", "content": state["query"]},
+                {"role": "assistant", "content": answer},
+            ],
+        }
+
+    async def save_memory(state: HealthButlerState) -> dict:
+        try:
+            if state.get("messages"):
+                await memory_mgr.save_conversation(
+                    state["user_id"],
+                    state["conversation_id"],
+                    state["messages"],
+                )
+        except Exception as e:
+            logger.warning("保存记忆失败: %s", e)
+        return {}
+
+    # ── 路由 ──
+
+    def route_after_knowledge(state: HealthButlerState) -> Literal["search_knowledge", "llm_answer"]:
+        kb_ctx = state.get("kb_context", {})
+        if kb_ctx.get("kb_loaded"):
+            return "search_knowledge"
+        return "llm_answer"
+
+    # ── 构建图 ──
+
+    builder.add_node("load_knowledge", load_knowledge)
+    builder.add_node("search_knowledge", search_knowledge)
+    builder.add_node("llm_answer", llm_answer)
+    builder.add_node("save_memory", save_memory)
+
+    builder.add_edge(START, "load_knowledge")
+    builder.add_conditional_edges("load_knowledge", route_after_knowledge)
+    builder.add_edge("search_knowledge", "llm_answer")
+    builder.add_edge("llm_answer", "save_memory")
+    builder.add_edge("save_memory", END)
+
+    checkpointer = MemorySaver()
+    graph = builder.compile(checkpointer=checkpointer)
+
+    return graph
+
+
+def _extract_health_keywords(query: str) -> list[str]:
+    """从 query 中提取健康相关关键词(菌属、指标、食物等名词)"""
+    # 简单策略: 提取2字以上的中文词作为候选
+    import re
+    # 提取中文字符组成的连续词段
+    chinese_words = re.findall(r'[\u4e00-\u9fff]{2,}', query)
+    # 提取英文/拉丁名词组
+    english_words = re.findall(r'[A-Za-z][A-Za-z\s]{2,20}[A-Za-z]', query)
+
+    keywords = []
+    # 过滤常见通用词
+    stopwords = {"请问", "帮我", "能不能", "可不可以", "怎么样", "为什么",
+                 "是什么", "怎么办", "什么意思", "好不好", "是否", "你好",
+                 "谢谢", "请", "我想", "需要", "关于", "一个", "这个",
+                 "那个", "哪些", "什么", "怎么", "如何", "多少", "是不是"}
+    for w in chinese_words:
+        if w not in stopwords and len(w) >= 2:
+            keywords.append(w)
+
+    for w in english_words:
+        w = w.strip()
+        if len(w) > 2:
+            keywords.append(w)
+
+    # 去重 + 限数量
+    seen = set()
+    deduped = []
+    for kw in keywords:
+        if kw not in seen:
+            seen.add(kw)
+            deduped.append(kw)
+
+    return deduped[:10]

+ 179 - 0
cfc-langgraph/app/graphs/health_coach_graph.py

@@ -0,0 +1,179 @@
+"""
+AI 健康教练 LangGraph - 专为儿童健康咨询设计的对话图
+Uses RAG retrieval from health knowledge base.
+"""
+from typing import TypedDict
+from langgraph.graph import StateGraph, START, END
+from langgraph.checkpoint.memory import MemorySaver
+from langchain_openai import ChatOpenAI, OpenAIEmbeddings
+from langchain_core.messages import SystemMessage, HumanMessage
+from app.rag.retriever import RagRetriever
+from app.memory.store import MemoryManager
+from app.config import settings
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class HealthCoachState(TypedDict):
+    query: str
+    user_id: int
+    child_id: int | None
+    conversation_id: str | None
+    context: dict | None
+    answer: str | None
+    sources: list[dict]
+    memory_messages: list | None
+
+
+HEALTH_COACH_PROMPT = """你是一个儿童健康教练,喜欢用提问的方式引导家长发现问题和解决方案。
+
+核心原则:
+1. 先问后答:不要直接给答案,用问题引导家长思考
+   - "孩子这种情况多久了?"、"平时饮食习惯是怎样的?"、"体重身高发育曲线有没有记录?"
+2. 用开放式问题了解情况,再给针对性建议
+3. 语气温暖,像一个耐心的教练在陪伴家长成长
+
+你的职责:
+1. 儿童体质调理(如挑食、瘦弱、易疲劳、反复生病)—— 用提问了解情况
+2. 科学喂养和营养建议 —— 先问饮食现状再给建议
+3. 日常保健(睡眠、运动、季节防护)—— 引导家长发现生活中的问题
+4. 体质辨识基础咨询 —— 通过提问帮助家长理解孩子体质特点
+5. 健康问题判断 —— 先收集症状信息再建议是否就医
+
+回答结构(教练风格):
+1. 提问:先问1-2个关键问题了解情况
+2. 分析:简短总结你观察到的问题
+3. 建议:给出2-3个可操作的建议
+4. 跟进问句:留一个后续跟进的问题,体现教练陪伴感
+
+原则:
+- 涉及严重症状(急性腹痛、高烧不退、严重外伤等)立刻建议就医,不提问
+- 不确定时明确说"建议咨询专业医生"
+- 始终用中文回答
+"""
+
+
+def create_health_coach_graph():
+    """创建AI健康教练 StateGraph"""
+    # LLM
+    llm = ChatOpenAI(
+        model=settings.llm_model,
+        api_key=settings.llm_api_key,
+        base_url=settings.llm_base_url,
+        temperature=settings.llm_temperature,
+    )
+
+    # 嵌入和检索
+    embeddings = OpenAIEmbeddings(
+        model=settings.embedding_model,
+        api_key=settings.effective_embedding_api_key,
+        base_url=settings.effective_embedding_base_url,
+    )
+    retriever = RagRetriever(collection_name="health_knowledge")
+    memory_mgr = MemoryManager(embeddings)
+
+    # 带工具的 LLM(健康知识库检索)
+    async def health_retrieve(query: str, user_id: int) -> list[dict]:
+        try:
+            results = await retriever.retrieve(
+                query,
+                filters={"user_id": user_id},
+                k=3,
+            )
+            return results
+        except Exception as e:
+            logger.warning("健康知识检索失败: %s", e)
+            return []
+
+    async def generate_answer(state: HealthCoachState) -> dict:
+        """带健康知识检索的 LLM 生成"""
+        messages = [SystemMessage(content=HEALTH_COACH_PROMPT)]
+
+        # 家庭上下文
+        ctx = state.get("context") or {}
+        if ctx:
+            child_info = []
+            if ctx.get("child_id"):
+                child_info.append(f"孩子ID: {ctx['child_id']}")
+            if ctx.get("family_id"):
+                child_info.append(f"家庭ID: {ctx['family_id']}")
+            if child_info:
+                messages.append(SystemMessage(
+                    content="用户背景信息: " + ", ".join(child_info)
+                ))
+
+        # 健康知识检索
+        retrieved = await health_retrieve(state["query"], state["user_id"])
+        sources = []
+        if retrieved:
+            context_texts = []
+            for i, r in enumerate(retrieved[:3]):
+                src = r.get("metadata", {})
+                title = src.get("title", "健康知识")
+                content = r["content"][:300]
+                context_texts.append(f"[{i+1}] {title}:\n{content}")
+                sources.append({
+                    "type": "knowledge",
+                    "title": title,
+                    "content": content,
+                })
+            knowledge_context = "\n\n---\n".join(context_texts)
+            messages.append(SystemMessage(
+                content=f"以下是与问题相关的健康知识参考:\n{knowledge_context}"
+            ))
+
+        # 长期记忆
+        try:
+            memories = await memory_mgr.recall(state["user_id"], state["query"])
+            if memories:
+                mem_text = "\n".join([f"- {m}" for m in memories[:3]])
+                messages.append(SystemMessage(
+                    content=f"该用户的历史健康咨询记录:\n{mem_text}"
+                ))
+        except Exception as e:
+            logger.warning("召回健康记忆失败: %s", e)
+
+        # 用户问题
+        messages.append(HumanMessage(content=state["query"]))
+
+        response = await llm.ainvoke(messages)
+        answer = response.content
+
+        return {
+            "answer": answer,
+            "sources": sources,
+            "memory_messages": [
+                {"role": "user", "content": state["query"]},
+                {"role": "assistant", "content": answer},
+            ],
+        }
+
+    async def save_memory(state: HealthCoachState) -> dict:
+        """保存对话到长期记忆"""
+        try:
+            if state.get("memory_messages"):
+                await memory_mgr.save_conversation(
+                    state["user_id"],
+                    state.get("conversation_id") or f"health_{state['user_id']}",
+                    state["memory_messages"],
+                )
+        except Exception as e:
+            logger.warning("保存健康教练记忆失败: %s", e)
+        return {}
+
+    # ── 构建图 ──
+
+    builder = StateGraph(HealthCoachState)
+
+    builder.add_node("generate_answer", generate_answer)
+    builder.add_node("save_memory", save_memory)
+
+    builder.add_edge(START, "generate_answer")
+    builder.add_edge("generate_answer", "save_memory")
+    builder.add_edge("save_memory", END)
+
+    checkpointer = MemorySaver()
+    graph = builder.compile(checkpointer=checkpointer)
+
+    return graph

+ 79 - 0
cfc-langgraph/app/graphs/recommend_graph.py

@@ -0,0 +1,79 @@
+from langchain_openai import ChatOpenAI
+from langchain_core.messages import SystemMessage, HumanMessage
+from app.tools.product_tools import search_product_by_keyword, search_activity_by_keyword, search_article_by_keyword
+from app.config import settings
+from app.monitoring import monitor_agent
+import json
+import logging
+
+logger = logging.getLogger(__name__)
+
+SYSTEM_PROMPT = """你是一个儿童成长营养推荐助手。根据用户的需求和营养标签, 推荐合适的商品、活动或文章。
+
+推荐原则:
+1. 首先尝试使用搜索工具查找匹配的内容
+2. 如果搜索结果为空, 基于你的知识给出建议
+3. 每项推荐必须附带推荐理由
+4. 以 JSON 格式输出推荐结果
+
+输出格式:
+{
+    "items": [
+        {
+            "source": "tool" 或 "knowledge",
+            "type": "product" / "activity" / "article",
+            "id": 数字,
+            "name": "名称",
+            "description": "描述",
+            "reason": "为什么推荐这个"
+        }
+    ]
+}
+"""
+
+
+class RecommendAgent:
+    def __init__(self):
+        self.llm = ChatOpenAI(
+            model=settings.llm_model,
+            api_key=settings.llm_api_key,
+            base_url=settings.llm_base_url,
+            temperature=settings.llm_temperature,
+        )
+        self.tools = [
+            search_product_by_keyword,
+            search_activity_by_keyword,
+            search_article_by_keyword,
+        ]
+        self.llm_with_tools = self.llm.bind_tools(self.tools)
+
+    @monitor_agent("recommend")
+    async def run(self, query: str, tags: list[str], limit: int = 5) -> dict:
+        """执行推荐 Agent, 返回推荐结果"""
+        # 如果传入了 tags, 构造搜索关键词
+        search_query = query or " ".join(tags)
+
+        messages = [
+            SystemMessage(content=SYSTEM_PROMPT),
+            HumanMessage(content=f"用户需求: {search_query}\n最大返回数量: {limit}\n请搜索并推荐合适的内容。"),
+        ]
+
+        # LangChain Tool calling 自动完成: LLM 决定调哪个 Tool → 工具返回结果 → LLM 组织回答
+        response = await self.llm_with_tools.ainvoke(messages)
+
+        # 尝试解析 JSON 输出
+        content = response.content
+        try:
+            # 提取 JSON 块
+            if "```json" in content:
+                json_str = content.split("```json")[1].split("```")[0].strip()
+            elif "```" in content:
+                json_str = content.split("```")[1].split("```")[0].strip()
+            else:
+                json_str = content.strip()
+            result = json.loads(json_str)
+            return result
+        except (json.JSONDecodeError, IndexError):
+            # 非 JSON 输出, 包装为文本回答
+            logger.warning("Agent 输出非 JSON, raw: %s", content[:200])
+            return {"items": [], "text": content}

+ 14 - 27
cfc-langgraph/app/main.py

@@ -3,43 +3,30 @@ import time
 import asyncio
 import logging
 from fastapi import FastAPI, Request
-from app.api import health
-# 以下模块文件被 f18dd86e 清空(0字节),恢复前注释避免 import 失败
-# from app.api import recommend, chat, analyze, tongue, adapter, report_parse, meal
+from fastapi.middleware.base import BaseHTTPMiddleware
+from app.api import health, recommend, chat, analyze, tongue, adapter, report_parse, meal, logs
 from app import monitoring
 from src.app import router as questionnaire_router
+from app.middleware import request_log_middleware
 
 logger = logging.getLogger(__name__)
 
 app = FastAPI(title="cfc-langgraph", version="0.3.0")
 
+# 注册请求日志中间件(必须在其他路由之前)
+app.add_middleware(BaseHTTPMiddleware, dispatch=request_log_middleware)
+
 app.include_router(health.router)
-# app.include_router(recommend.router)
-# app.include_router(chat.router)
-# app.include_router(analyze.router)
-# app.include_router(tongue.router)
-# app.include_router(adapter.router)
+app.include_router(recommend.router)
+app.include_router(chat.router)
+app.include_router(analyze.router)
+app.include_router(tongue.router)
+app.include_router(adapter.router)
 app.include_router(monitoring.router)
 app.include_router(questionnaire_router)
-# app.include_router(report_parse.router)
-# app.include_router(meal.router)
-
-
-@app.middleware("http")
-async def timing_middleware(request: Request, call_next):
-    start = time.perf_counter()
-    response = await call_next(request)
-    elapsed = time.perf_counter() - start
-
-    if elapsed > 5:
-        logger.warning("SLOW_REQUEST: %s %s took %.2fs",
-                       request.method, request.url.path, elapsed)
-    else:
-        logger.debug("REQUEST: %s %s took %.2fs",
-                     request.method, request.url.path, elapsed)
-
-    response.headers["X-Response-Time"] = f"{elapsed:.3f}s"
-    return response
+app.include_router(report_parse.router)
+app.include_router(meal.router)
+app.include_router(logs.router)
 
 
 @app.on_event("startup")

+ 99 - 0
cfc-langgraph/app/middleware.py

@@ -0,0 +1,99 @@
+"""请求日志中间件 — 记录最近 MAX_ENTRIES 条请求到内存"""
+import time
+import asyncio
+from collections import deque
+from typing import Optional
+from fastapi import Request, Response
+import logging
+
+logger = logging.getLogger(__name__)
+
+MAX_ENTRIES = 500
+_request_log: deque = deque(maxlen=MAX_ENTRIES)
+
+
+def get_request_log() -> list:
+    """返回请求日志副本(最近在前)"""
+    return list(_request_log)[::-1]
+
+
+def clear_request_log() -> int:
+    """清空日志,返回清空条数"""
+    n = len(_request_log)
+    _request_log.clear()
+    return n
+
+
+async def request_log_middleware(request: Request, call_next) -> Response:
+    start_time = time.perf_counter()
+    client_ip = request.client.host if request.client else "-"
+    
+    # 读取请求体(不重复消费)
+    body = None
+    if request.method in ("POST", "PUT", "PATCH"):
+        try:
+            raw = await request.body()
+            if raw:
+                try:
+                    body = raw.decode("utf-8")
+                    if len(body) > 2000:
+                        body = body[:2000] + "...(truncated)"
+                except Exception:
+                    body = f"<binary {len(raw)} bytes>"
+        except Exception:
+            pass
+    
+    try:
+        response = await call_next(request)
+    except Exception as exc:
+        elapsed = time.perf_counter() - start_time
+        status_code = 500
+        response_body = ""
+        logger.error("REQUEST EXCEPTION: %s %s %.2fs error=%r",
+                     request.method, request.url.path, elapsed, exc)
+        _request_log.appendleft({
+            "ts": time.strftime("%H:%M:%S"),
+            "method": request.method,
+            "path": request.url.path,
+            "status": 500,
+            "duration_ms": round(elapsed * 1000),
+            "client_ip": client_ip,
+            "body": body,
+            "error": str(exc)[:500],
+        })
+        # 构造错误响应
+        from fastapi.responses import JSONResponse
+        return JSONResponse(status_code=500, content={"detail": str(exc)})
+    
+    elapsed = time.perf_counter() - start_time
+    status_code = response.status_code
+    
+    # 读取响应体(不可逆,只取内容长度)
+    response_body = ""
+    if hasattr(response, "body") and response.body:
+        try:
+            response_body = response.body.decode("utf-8")[:1000]
+        except Exception:
+            response_body = f"<binary {len(response.body)} bytes>"
+    
+    # 写入日志
+    entry = {
+        "ts": time.strftime("%H:%M:%S"),
+        "method": request.method,
+        "path": request.url.path,
+        "status": status_code,
+        "duration_ms": round(elapsed * 1000),
+        "client_ip": client_ip,
+        "body": body,
+        "response": response_body,
+    }
+    _request_log.appendleft(entry)
+    
+    # 超5秒标记 warning
+    if elapsed > 5:
+        logger.warning("SLOW_REQUEST: %s %s %.2fs", request.method, request.url.path, elapsed)
+    else:
+        logger.debug("REQUEST: %s %s %.2fs", request.method, request.url.path, elapsed)
+    
+    response.headers["X-Response-Time"] = f"{elapsed:.3f}s"
+    return response

+ 3 - 0
cfc-langgraph/app/models/__init__.py

@@ -0,0 +1,3 @@
+from .common import UserContext, SourceInfo
+from .chat import ChatRequest, ChatResponse
+from .recommend import RecommendRequest, RecommendResponse, RecommendItem

+ 18 - 0
cfc-langgraph/app/models/chat.py

@@ -0,0 +1,18 @@
+from pydantic import BaseModel
+from typing import Optional
+from .common import UserContext, SourceInfo
+
+
+class ChatRequest(BaseModel):
+    query: str
+    user_id: int
+    conversation_id: str = ""
+    context: Optional[UserContext] = None
+
+
+class ChatResponse(BaseModel):
+    answer: str
+    conversation_id: str
+    sources: list[SourceInfo] = []
+    tasks: list[dict] = []
+    trace_id: str = ""

+ 18 - 0
cfc-langgraph/app/models/common.py

@@ -0,0 +1,18 @@
+from pydantic import BaseModel
+from typing import Optional
+
+
+class UserContext(BaseModel):
+    """Java 侧传来的业务上下文"""
+    child_id: Optional[int] = None
+    report_id: Optional[int] = None
+    survey_id: Optional[int] = None
+    family_id: Optional[int] = None
+    mascot_code: Optional[str] = None
+
+
+class SourceInfo(BaseModel):
+    """回答引用来源"""
+    type: str  # knowledge / tool
+    title: str
+    score: Optional[float] = None

+ 48 - 0
cfc-langgraph/app/models/meal.py

@@ -0,0 +1,48 @@
+from pydantic import BaseModel
+from typing import Optional, List
+
+
+class FoodRecognizeRequest(BaseModel):
+    image_url: str
+    prompt: str = "识别图片中的主要食材,返回 JSON 格式:[{name, confidence, category}]"
+
+
+class FoodsItem(BaseModel):
+    name: str
+    confidence: float
+    category: str = "unknown"
+
+
+class FoodRecognizeResponse(BaseModel):
+    foods: List[FoodsItem]
+    raw_response: str = ""
+    trace_id: str = ""
+
+
+class MenuGenerateRequest(BaseModel):
+    selected_foods: str
+    participants: str
+    date: str
+    allergies: Optional[str] = None
+    avoid_foods: Optional[str] = None
+    health_goals: Optional[str] = None
+    spice_level: Optional[int] = None
+    cuisine_pref: Optional[str] = None
+
+
+class MenuIngredient(BaseModel):
+    name: str
+    grams: int
+
+
+class MenuDish(BaseModel):
+    name: str
+    ingredients: List[MenuIngredient]
+    cooking_method: str = ""
+    nutrition: Optional[dict] = None
+    notes: str = ""
+
+
+class MenuGenerateResponse(BaseModel):
+    menu_json: str = ""
+    trace_id: str = ""

+ 29 - 0
cfc-langgraph/app/models/recommend.py

@@ -0,0 +1,29 @@
+from pydantic import BaseModel
+from typing import Optional
+
+
+class RecommendItem(BaseModel):
+    type: str  # product / activity / article
+    id: int
+    name: str
+    description: str = ""
+    cover_image: str = ""
+    price: Optional[float] = None
+    url: str = ""
+    score: float = 0.0
+    reason: str = ""
+
+
+class RecommendRequest(BaseModel):
+    user_id: int
+    query: str = ""
+    tags: list[str] = []
+    types: Optional[list[str]] = None
+    limit: int = 5
+    context: Optional[dict] = None
+
+
+class RecommendResponse(BaseModel):
+    items: list[RecommendItem]
+    source: str = ""  # sql / vector / hybrid
+    trace_id: str = ""

+ 6 - 4
cfc-langgraph/app/monitoring.py

@@ -124,14 +124,16 @@ async def detailed_health():
         from app.tools.java_client import JavaClient
         client = JavaClient()
         jc = await client._get_client()
-        resp = await jc.get("/health")
-        if resp.status_code == 200:
+        resp = await jc.post("/api/admin-auth/login-by-password", json={})
+        if resp.status_code == 401:
+            status["components"]["java_backend"] = {"status": "ok", "detail": "需认证(正常)"}
+        elif resp.status_code == 200:
             status["components"]["java_backend"] = {"status": "ok"}
         else:
-            status["components"]["java_backend"] = {"status": "error", "code": resp.status_code}
+            status["components"]["java_backend"] = {"status": "error", "code": resp.status_code, "detail": f"HTTP {resp.status_code}"}
             status["status"] = "degraded"
     except Exception as e:
-        status["components"]["java_backend"] = {"status": "error", "message": str(e)}
+        status["components"]["java_backend"] = {"status": "error", "detail": str(e)}
         status["status"] = "degraded"
 
     return status

+ 166 - 0
cfc-langgraph/app/parsers/pdf_parser.py

@@ -0,0 +1,166 @@
+"""
+菌群报告 PDF 解析器
+从 extract_full_report_v5.py 提取核心逻辑,封装为可调用函数
+"""
+import re
+from PyPDF2 import PdfReader
+
+# === 常量 ===
+RADICAL_MAP = {
+    '\u2f52': '\u6c0f', '\u2f51': '\u6bcd', '\u2f59': '\u6b6f',
+    '\u2f04': '\u4e59', '\u2f20': '\u4e00', '\u2f21': '\u4e28',
+    '\u2f22': '\u4e3f', '\u2f23': '\u4e39', '\u2f2b': '\u5c38',
+    '\u2f2d': '\u5c71', '\u2f44': '\u4e59', '\u2f53': '\u6c14',
+    '\u2f55': '\u6c34', '\u2f5c': '\u725b', '\u2f5f': '\u7389',
+    '\u2f7a': '\u7f8a', '\u2f81': '\u8089', '\u2f83': '\u81ea',
+    '\u2f8a': '\u8272', '\u2f8e': '\u8840', '\u2f95': '\u79be',
+    '\u2f96': '\u8c46', '\u2faf': '\u9762', '\u2fb9': '\u9999',
+    '\u2fca': '\u9ed1', '\u2ec9': '\u8d1d', '\u2edd': '\u98df',
+    '\u2ee2': '\u9a6c', '\u2ee9': '\u9ec4',
+}
+
+KNOWN_MACRO = ['碳水化合物', '蛋白质', '脂肪', '纤维素', '乳制品']
+KNOWN_AMINO = ['苏氨酸', '异亮氨酸', '亮氨酸', '赖氨酸', '蛋氨酸', '胱氨酸',
+    '苯丙氨酸', '酪氨酸', '缬氨酸', '组氨酸', '丙氨酸', '丝氨酸', '甘氨酸',
+    '脯氨酸', '谷氨酸', '天门冬氨酸', '天冬氨酸', '天冬酰胺', '谷氨酰胺',
+    '精氨酸', '色氨酸']
+KNOWN_VITAMINS = ['维生素A', '维生素B1', '维生素B2', '维生素B5', '维生素B6',
+    '叶酸', '维生素B12', '维生素C', '维生素D', '维生素K2', '维生素E']
+KNOWN_TRACE = ['铁', '锌']
+KNOWN_DISEASE_RISKS = ['炎症性肠炎', '肠易激综合征', '感染性腹泻', '自闭症',
+    '抑郁症', '甲状腺疾病', '肺部感染或疾病', '自体免疫病', '结直肠癌',
+    '肥胖', '便秘', '过敏', '失眠', '肝病', '肾病', '胃病', '胆病',
+    '心脑血管疾病', 'II型糖尿病']
+KNOWN_BARRIER = ['肠道炎症水平', '肠道产气', '肠道屏障', '脂多糖LPS',
+    '次级胆汁酸', '对甲酚(p-Cresol)', '吲哚', '苯酚', '腐胺', '硫化氢', '尸胺']
+KNOWN_SCFA = ['丁酸盐(Butyrate)', '丙酸盐(Propionate)', '乙酸盐(Acetate)', '异戊酸盐(Isovaleric)']
+KNOWN_NEURO = ['血清素(5-HT)', 'γ-氨基丁酸(GABA)', '谷氨酸(Glutamate)',
+    '色氨酸(Tryptophan)', 'DOPAC', '多巴胺', '组胺(Histamine)', '一氧化氮',
+    '喹啉(Quinolinic)', '维生素K2', '肌醇(Inositol)', '肾上腺素',
+    '去甲肾上腺素', '乙酰胆碱', '皮质醇']
+KNOWN_ANTIBIOTICS = ['β-内酰胺酶类', '氨基糖苷类', '大环内酯类', '呋喃类',
+    '喹诺酮类', '磺胺类', '甲氧苄啶类', '氯霉素类', '四环素类']
+KNOWN_PATHOGENS = ['幽门螺杆菌', '艰难梭菌', '沙门氏菌', '志贺氏菌', '弯曲杆菌']
+
+
+def norm(s):
+    return ''.join(RADICAL_MAP.get(c, c) for c in s)
+
+
+def detect_format(lines):
+    """检测 triplet / inline 格式"""
+    text = '\n'.join(lines)
+    has_triplet = '指标范围' in text and '疾病风险评估' in text
+    for line in lines:
+        if len(line) > 15 and re.search(r'[\u4e00-\u9fff]+[\d.]+[\u4e00-\u9fff/]+', line):
+            for known in KNOWN_DISEASE_RISKS:
+                if known in line:
+                    return 'inline'
+    return 'triplet' if has_triplet else 'inline'
+
+
+def extract_text(file_path):
+    """读取 PDF 并提取文本"""
+    reader = PdfReader(file_path)
+    lines = []
+    for page in reader.pages:
+        text = norm(page.extract_text() or '')
+        for line in text.split('\n'):
+            ls = line.strip()
+            if ls:
+                lines.append(ls)
+    return lines
+
+
+def parse_overview(lines):
+    """提取报告概述"""
+    text = '\n'.join(lines)
+    r = {}
+    m = re.search(r'编号[::\s]*(\d+)', text)
+    if m: r['report_number'] = m.group(1)
+    m = re.search(r'姓名[::\s]*([\u4e00-\u9fff]{2,10})', text)
+    if m: r['person_name'] = re.sub(r'(编号|年龄|性别|备注|肠道).*', '', m.group(1))[:4]
+    m = re.search(r'年龄[::\s]*(\d+)', text)
+    if m: r['age'] = int(m.group(1))
+    m = re.search(r'性别[::\s]*([\u4e00-\u9fff])', text)
+    if m: r['gender'] = 'male' if m.group(1) == '男' else 'female'
+    for kw in ['健康总分', '菌群健康', '慢病控制', '营养均衡', '肠道菌群平衡',
+               '菌群多样性', '有益菌', '有害菌', '核心菌属']:
+        m = re.search(rf'{kw}\s*(\d+)', text)
+        if m: r[kw] = int(m.group(1))
+    m = re.search(r'肠道预测年龄[::\s]*([\d.]+)', text)
+    if m: r['gut_age'] = m.group(1)
+    m = re.search(r'肠型[::\s]*(\S+)', text)
+    if m: r['gut_type'] = m.group(1)
+    return r
+
+
+def parse_triplet_until(lines, stop_markers):
+    """三元组解析:3行一组 名称/数值/状态"""
+    results = []
+    i = 0
+    while i < len(lines):
+        if any(lines[i] == sm or lines[i].startswith(sm) for sm in stop_markers):
+            break
+        if lines[i] in ('指标范围', '名称', '丰度', '评估'):
+            i += 1
+            continue
+        name = lines[i]
+        if i + 2 >= len(lines): break
+        val = lines[i + 1]
+        status = lines[i + 2]
+        if re.match(r'^-?\d+\.?\d*$', val):
+            results.append({'name': name, 'value': val, 'status': status})
+            i += 3
+        else:
+            i += 1
+    return results
+
+
+def parse_report_pdf(file_path: str) -> dict:
+    """主函数:解析 PDF 返回结构化数据"""
+    lines = extract_text(file_path)
+    fmt = detect_format(lines)
+    result = {'format': fmt, 'overview': parse_overview(lines)}
+
+    # 疾病风险评估
+    for i, line in enumerate(lines):
+        if '疾病风险评估' in line and '注' not in line:
+            risks = parse_triplet_until(lines[i+1:],
+                ['主要营养评估', '氨基酸评估', '维生素评估', '微量元素评估', '抗生素风险评估'])
+            result['disease_risks'] = [r for r in risks if '注' not in r['name']]
+            break
+
+    # 主要营养评估
+    for i, line in enumerate(lines):
+        if '主要营养评估' in line:
+            nutrients = parse_triplet_until(lines[i+1:], ['氨基酸评估'])
+            result['nutrition'] = nutrients[:5]
+            break
+
+    # 氨基酸评估
+    for i, line in enumerate(lines):
+        if '氨基酸评估' in line:
+            aminos = parse_triplet_until(lines[i+1:], ['维生素评估', '微量元素评估'])
+            result['amino_acids'] = aminos
+            break
+
+    # 维生素评估
+    for i, line in enumerate(lines):
+        if '维生素评估' in line:
+            vits = parse_triplet_until(lines[i+1:], ['微量元素评估', '抗生素风险评估'])
+            result['vitamins'] = [r for r in vits if '维生素' in r['name']]
+            result['trace_elements'] = [r for r in vits if '维生素' not in r['name']]
+            break
+
+    return result
+
+
+def parse_report_pdf_with_fallback(file_path: str) -> dict:
+    """算法解析 + 简单校验,返回结构化数据"""
+    result = parse_report_pdf(file_path)
+    # 简单校验:如果关键字段缺失,标记为解析不完整
+    if not result.get('overview', {}).get('overallScore') and \
+       not result.get('overview', {}).get('健康总分'):
+        result['_parse_incomplete'] = True
+    return result

+ 77 - 0
cfc-langgraph/app/tasks/knowledge_sync.py

@@ -0,0 +1,77 @@
+"""知识库同步定时任务: 定期从 Java 拉取文章, 更新 ChromaDB"""
+from app.rag.loader import KnowledgeLoader
+from app.rag.splitter import get_knowledge_splitter
+from app.rag.embeddings import get_embeddings
+from app.config import settings
+from langchain_chroma import Chroma
+from langchain_core.documents import Document
+import logging
+import os
+import json
+
+logger = logging.getLogger(__name__)
+
+SYNC_STATE_FILE = os.path.join(settings.chroma_db_path, ".sync_state")
+
+
+def _load_sync_state() -> dict:
+    try:
+        if os.path.exists(SYNC_STATE_FILE):
+            with open(SYNC_STATE_FILE) as f:
+                return json.load(f)
+    except Exception:
+        pass
+    return {"last_sync": "2000-01-01T00:00:00"}
+
+
+def _save_sync_state(state: dict):
+    os.makedirs(os.path.dirname(SYNC_STATE_FILE), exist_ok=True)
+    with open(SYNC_STATE_FILE, "w") as f:
+        json.dump(state, f)
+
+
+async def sync_knowledge_base():
+    """执行知识库同步 (全量+增量)"""
+    loader = KnowledgeLoader()
+    splitter = get_knowledge_splitter()
+    embeddings = get_embeddings()
+
+    vectorstore = Chroma(
+        collection_name="cfc_knowledge",
+        embedding_function=embeddings,
+        persist_directory=settings.chroma_db_path,
+    )
+
+    state = _load_sync_state()
+
+    articles = await loader.load_updated_since(state["last_sync"])
+    if not articles:
+        logger.info("知识库同步: 无更新内容")
+        return
+
+    docs = loader.format_for_indexing(articles)
+
+    chunks = []
+    for doc in docs:
+        split_texts = splitter.split_text(doc["content"])
+        for i, text in enumerate(split_texts):
+            metadata = dict(doc["metadata"])
+            metadata["chunk_index"] = i
+            chunks.append(Document(page_content=text, metadata=metadata))
+
+    if not chunks:
+        logger.info("知识库同步: 无新增块")
+        return
+
+    BATCH_SIZE = 100
+    for i in range(0, len(chunks), BATCH_SIZE):
+        batch = chunks[i:i + BATCH_SIZE]
+        await vectorstore.aadd_documents(batch)
+        vectorstore.persist()
+        logger.debug("知识库同步进度: %d/%d", i + len(batch), len(chunks))
+
+    import datetime
+    state["last_sync"] = datetime.datetime.now().isoformat()
+    _save_sync_state(state)
+
+    logger.info("知识库同步完成: 新增 %d 篇文章, %d 个块", len(articles), len(chunks))

+ 88 - 0
cfc-langgraph/app/tools/java_client.py

@@ -0,0 +1,88 @@
+import httpx
+from typing import Optional
+from app.config import settings
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class JavaClient:
+    """Java 后端 HTTP 客户端 (单例, 复用连接池)"""
+
+    _instance: Optional["JavaClient"] = None
+
+    def __new__(cls):
+        if cls._instance is None:
+            cls._instance = super().__new__(cls)
+            cls._instance._client = None
+            cls._instance._base_url = settings.java_base_url
+        return cls._instance
+
+    async def _get_client(self) -> httpx.AsyncClient:
+        if self._client is None:
+            limits = httpx.Limits(
+                max_connections=10,
+                max_keepalive_connections=5,
+                keepalive_expiry=30,
+            )
+            self._client = httpx.AsyncClient(
+                base_url=self._base_url,
+                timeout=httpx.Timeout(10.0, connect=3.0),
+                limits=limits,
+            )
+        return self._client
+
+    async def close(self):
+        if self._client:
+            await self._client.aclose()
+            self._client = None
+            JavaClient._instance = None
+
+    async def get_published_articles(self) -> list[dict]:
+        client = await self._get_client()
+        resp = await client.post("/api/article/list", json={"status": "published", "limit": 1000})
+        data = resp.json()
+        if data.get("code") == 200:
+            return data.get("data", [])
+        return []
+
+    async def search_products(self, keyword: str, limit: int = 5) -> list[dict]:
+        client = await self._get_client()
+        resp = await client.post("/api/product/search", json={
+            "keyword": keyword, "status": "上架", "limit": limit
+        })
+        data = resp.json()
+        if data.get("code") == 200:
+            return data.get("data", [])
+        return []
+
+    async def search_activities(self, keyword: str, limit: int = 5) -> list[dict]:
+        client = await self._get_client()
+        resp = await client.post("/api/activity/search", json={
+            "keyword": keyword, "status": "published", "limit": limit
+        })
+        data = resp.json()
+        if data.get("code") == 200:
+            return data.get("data", [])
+        return []
+
+    async def search_articles(self, keyword: str, limit: int = 5) -> list[dict]:
+        client = await self._get_client()
+        resp = await client.post("/api/article/search", json={
+            "keyword": keyword, "status": "published", "limit": limit
+        })
+        data = resp.json()
+        if data.get("code") == 200:
+            return data.get("data", [])
+        return []
+
+    async def get_user_context(self, user_id: int, params: Optional[dict] = None) -> dict:
+        client = await self._get_client()
+        resp = await client.post(settings.effective_java_context_url, json={
+            "user_id": str(user_id),
+            "params": params or {},
+        })
+        data = resp.json()
+        if data.get("code") == 200:
+            return data.get("data", {})
+        return {}

+ 58 - 0
cfc-langgraph/app/tools/product_tools.py

@@ -0,0 +1,58 @@
+from langchain_core.tools import tool
+from app.tools.java_client import JavaClient
+import logging
+
+logger = logging.getLogger(__name__)
+_java = JavaClient()
+
+
+@tool
+async def search_product_by_keyword(keyword: str, limit: int = 5) -> str:
+    """按关键词搜索上架商品, 返回 JSON 商品列表"""
+    try:
+        products = await _java.search_products(keyword, limit)
+        if not products:
+            return "[]"
+        return str([{
+            "id": p["id"],
+            "name": p["name"],
+            "price": p.get("price"),
+            "description": p.get("intro") or p.get("description", ""),
+        } for p in products])
+    except Exception as e:
+        logger.warning("搜索商品失败: %s", e)
+        return "[]"
+
+
+@tool
+async def search_activity_by_keyword(keyword: str, limit: int = 5) -> str:
+    """按关键词搜索进行中的活动, 返回 JSON 活动列表"""
+    try:
+        activities = await _java.search_activities(keyword, limit)
+        if not activities:
+            return "[]"
+        return str([{
+            "id": a["id"],
+            "name": a["title"],
+            "description": a.get("description", ""),
+        } for a in activities])
+    except Exception as e:
+        logger.warning("搜索活动失败: %s", e)
+        return "[]"
+
+
+@tool
+async def search_article_by_keyword(keyword: str, limit: int = 5) -> str:
+    """按关键词搜索已发布文章, 返回 JSON 文章列表"""
+    try:
+        articles = await _java.search_articles(keyword, limit)
+        if not articles:
+            return "[]"
+        return str([{
+            "id": a["id"],
+            "name": a["title"],
+            "summary": a.get("summary", ""),
+        } for a in articles])
+    except Exception as e:
+        logger.warning("搜索文章失败: %s", e)
+        return "[]"

+ 63 - 0
cfc-langgraph/app/tools/report_tools.py

@@ -0,0 +1,63 @@
+from langchain_core.tools import tool
+from app.tools.java_client import JavaClient
+import logging
+
+logger = logging.getLogger(__name__)
+_java = JavaClient()
+
+
+@tool
+async def get_report_detail(report_id: int) -> str:
+    """获取健康报告的完整详情, 包含各维度评分和解读"""
+    try:
+        client = _java
+        resp = await client._get_client().post(
+            "/api/health/report/detail",
+            json={"reportId": report_id},
+        )
+        data = resp.json()
+        if data.get("code") == 200:
+            import json
+            return json.dumps(data.get("data", {}), ensure_ascii=False)
+        return "{}"
+    except Exception as e:
+        logger.warning("获取报告详情失败: %s", e)
+        return "{}"
+
+
+@tool
+async def get_survey_data(report_id: int) -> str:
+    """获取与报告关联的调研问卷数据"""
+    try:
+        client = _java
+        resp = await client._get_client().post(
+            "/api/survey/completed",
+            json={"reportId": report_id},
+        )
+        data = resp.json()
+        if data.get("code") == 200:
+            import json
+            return json.dumps(data.get("data", {}), ensure_ascii=False)
+        return "{}"
+    except Exception as e:
+        logger.warning("获取问卷数据失败: %s", e)
+        return "{}"
+
+
+@tool
+async def get_dimension_scores(family_id: int) -> str:
+    """获取家庭成员的五维能量分数 (身/心/智/行/富)"""
+    try:
+        client = _java
+        resp = await client._get_client().post(
+            "/api/dimension/scores",
+            json={"familyId": family_id},
+        )
+        data = resp.json()
+        if data.get("code") == 200:
+            import json
+            return json.dumps(data.get("data", {}), ensure_ascii=False)
+        return "{}"
+    except Exception as e:
+        logger.warning("获取维度分数失败: %s", e)
+        return "{}"

+ 54 - 0
cfc-langgraph/deploy.sh

@@ -0,0 +1,54 @@
+#!/bin/bash
+# LangGraph 部署脚本
+# 用法: ./deploy.sh [生产|测试]
+# 会将本地 cfc-langgraph/ 代码同步到远程服务器并重启服务
+#
+# 远程服务器: ai.etotem.com.cn (111.228.6.214)
+# 部署路径: /opt/sc-data/cfc/cfc-langgraph/
+
+set -e
+
+ENV="${1:-prod}"
+REMOTE_USER="root"
+REMOTE_HOST="111.228.6.214"
+REMOTE_PATH="/opt/sc-data/cfc/cfc-langgraph"
+
+SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
+
+echo "=== LangGraph Deploy ($ENV) ==="
+echo "本地: $SCRIPT_DIR"
+echo "远程: $REMOTE_USER@$REMOTE_HOST:$REMOTE_PATH"
+
+# 同步代码(排除虚拟环境和缓存)
+echo "[1/3] 同步代码..."
+rsync -av --delete \
+  --exclude='.venv' \
+  --exclude='__pycache__' \
+  --exclude='*.pyc' \
+  --exclude='.env' \
+  --exclude='.env.production' \
+  --exclude='data/' \
+  --exclude='node_modules' \
+  --exclude='.git/' \
+  "$SCRIPT_DIR/" \
+  "${REMOTE_USER}@${REMOTE_HOST}:${REMOTE_PATH}/"
+
+# 重启服务
+echo "[2/3] 重启 LangGraph 服务..."
+ssh "${REMOTE_USER}@${REMOTE_HOST}" "
+  cd ${REMOTE_PATH} && \
+  docker-compose restart langgraph-svc && \
+  sleep 3 && \
+  docker-compose ps langgraph-svc
+"
+
+# 健康检查
+echo "[3/3] 健康检查..."
+sleep 5
+if curl -sf "https://ai.etotem.com.cn/health" > /dev/null; then
+  echo "✅ 部署成功!服务地址: https://ai.etotem.com.cn"
+  curl -s "https://ai.etotem.com.cn/api/v1/health" | python3 -m json.tool 2>/dev/null || true
+else
+  echo "❌ 健康检查失败,请查看日志: ssh ${REMOTE_USER}@${REMOTE_HOST} 'docker-compose -f ${REMOTE_PATH}/docker-compose.yml logs langgraph-svc --tail=50'"
+  exit 1
+fi

+ 68 - 0
cfc-langgraph/docker-compose.yml

@@ -0,0 +1,68 @@
+version: "3.8"
+
+services:
+  cfc-backend:
+    build:
+      context: ../cfc-backend
+      dockerfile: Dockerfile
+    ports:
+      - "9082:9082"
+    environment:
+      - SPRING_PROFILES_ACTIVE=prod
+      - python.enabled=true
+      - python.base-url=http://langgraph-svc:9000
+    networks:
+      - cfc-net
+    restart: unless-stopped
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:9082/actuator/health"]
+      interval: 30s
+      timeout: 5s
+      retries: 3
+
+  langgraph-svc:
+    build:
+      context: .
+      dockerfile: Dockerfile
+    ports:
+      - "9000:9000"
+    env_file:
+      - .env.production
+    environment:
+      - JAVA_BASE_URL=http://172.17.0.1:9082
+      - JAVA_CONTEXT_URL=http://172.17.0.1:9082/api/ai/context
+      - CHROMA_DB_PATH=/data/chroma_db
+      - JSON_LOGS=true
+      # LangSmith 三个变量从 env_file (.env.production) 读取,不再覆盖
+    volumes:
+      # 共享 Java 文件上传目录(LangGraph 读取已上传的 PDF 文件)
+      - /data/cfc-uploads:/data/cfc-uploads
+      # Chroma 向量数据库持久化
+      - langgraph-data:/data/chroma_db
+    networks:
+      - cfc-net
+    restart: unless-stopped
+    healthcheck:
+      test: ["CMD", "python", "-c", "import urllib.request,sys; sys.exit(0 if urllib.request.urlopen('http://localhost:9000/health', timeout=5).status==200 else 1)"]
+      interval: 30s
+      timeout: 5s
+      retries: 3
+
+
+  prometheus:
+    image: prom/prometheus:latest
+    volumes:
+      - ./prometheus.yml:/etc/prometheus/prometheus.yml
+      - prometheus-data:/prometheus
+    networks:
+      - cfc-net
+    restart: unless-stopped
+    profiles:
+      - monitoring
+
+networks:
+  cfc-net:
+
+volumes:
+  langgraph-data:
+  prometheus-data:

+ 97 - 0
cfc-langgraph/docs/deployment.md

@@ -0,0 +1,97 @@
+# LangGraph Sidecar 部署文档
+
+## 前置依赖
+
+- Docker + Docker Compose (推荐)
+- 或 Python 3.11+ + pip (直接部署)
+- LLM API Key (DeepSeek / OpenAI 兼容)
+
+## 目录结构
+
+```
+cfc-langgraph/
+├── Dockerfile
+├── docker-compose.yml
+├── prometheus.yml
+├── .env.production          # 生产配置 (手动创建, 不上传 git)
+├── app/                     # 服务代码
+│   ├── main.py              # FastAPI 入口
+│   ├── config.py            # 配置管理
+│   ├── api/                 # HTTP 接口
+│   ├── agents/              # Agent 定义
+│   ├── graphs/              # LangGraph StateGraph
+│   ├── tools/               # Agent Tool
+│   ├── rag/                 # RAG Pipeline
+│   ├── memory/              # 三层记忆
+│   └── tasks/               # 定时任务
+├── data/
+│   └── chroma_db/           # ChromaDB 持久化 (自动创建)
+└── docs/
+    ├── deployment.md        # 本文件
+    └── operations.md        # 运维手册
+```
+
+## Docker Compose 部署
+
+```bash
+# 1. 创建生产配置
+cp .env.example .env.production
+# 编辑 .env.production 填入 LLM_API_KEY
+
+# 2. 构建并启动
+docker-compose up -d
+
+# 3. 验证
+curl http://localhost:9000/api/v1/health
+
+# 4. 查看日志
+docker-compose logs -f langgraph-svc
+```
+
+## 直接部署 (无 Docker)
+
+```bash
+# 1. 安装依赖
+pip install -e .
+
+# 2. 配置环境变量
+export LLM_API_KEY=sk-xxx
+export JAVA_BASE_URL=http://localhost:9082
+
+# 3. 启动
+gunicorn app.main:app \
+  --worker-class uvicorn.workers.UvicornWorker \
+  --bind 0.0.0.0:9000 \
+  --workers 2 \
+  --timeout 60 \
+  --access-logfile - \
+  --error-logfile -
+```
+
+## 环境变量说明
+
+| 变量 | 必填 | 说明 |
+|------|------|------|
+| LLM_API_KEY | 是 | LLM API Key |
+| LLM_BASE_URL | 否 | 默认 https://api.deepseek.com/v1 |
+| LLM_MODEL | 否 | 默认 deepseek-chat |
+| JAVA_BASE_URL | 是 | Java 后端地址 |
+| CHROMA_DB_PATH | 否 | ChromaDB 持久化路径, 默认 ./data/chroma_db |
+| LOG_LEVEL | 否 | 日志级别, 默认 info |
+| JSON_LOGS | 否 | 启用 JSON 日志格式, 默认 false |
+
+## 健康检查
+
+```
+GET /api/v1/health
+
+Response:
+{
+  "status": "ok",
+  "components": {
+    "chromadb": {"status": "ok", "path": "...", "exists": true},
+    "llm": {"status": "ok"},
+    "java_backend": {"status": "ok"}
+  }
+}
+```

+ 121 - 0
cfc-langgraph/docs/operations.md

@@ -0,0 +1,121 @@
+# LangGraph Sidecar 运维手册
+
+## 日常监控
+
+### 1. 健康检查
+
+生产环境建议配置 30 秒定时健康检查:
+
+```bash
+curl -f http://localhost:9000/api/v1/health
+```
+
+预期返回 `{"status":"ok"}`。组件降级时返回 `{"status":"degraded"}`。
+
+### 2. Prometheus 指标
+
+```
+GET /metrics
+
+关键指标:
+- agent_calls_total{agent_type="chat|recommend|analysis"}
+- agent_duration_seconds{agent_type="..."}
+- llm_calls_total{model="deepseek-chat",status="ok|error"}
+- llm_duration_seconds{model="..."}
+- kb_sync_duration_seconds        # 知识库同步耗时
+```
+
+### 3. 日志
+
+JSON 格式日志可直接接入 ELK / Loki:
+
+```json
+{"timestamp":"2026-07-20T10:00:00","level":"WARN","logger":"app.main","message":"SLOW_REQUEST: POST /api/v1/chat took 8.23s"}
+```
+
+## 常见问题
+
+### Python 服务无法启动
+
+```bash
+# 检查依赖
+pip list | grep -E "fastapi|langchain|langgraph"
+
+# 检查配置
+python -c "from app.config import settings; print(settings.llm_model)"
+
+# 检查端口占用
+netstat -ano | grep 9000
+```
+
+### LLM 调用失败
+
+1. 检查 `.env.production` 中的 `LLM_API_KEY` 是否有效
+2. 检查 `LLM_BASE_URL` 是否可访问
+3. 查看日志: `docker-compose logs langgraph-svc | grep llm_call`
+
+### ChromaDB 损坏
+
+```bash
+# 删除后重建 (知识库会自动同步)
+rm -rf data/chroma_db/
+docker-compose restart langgraph-svc
+
+# 或触发手动同步
+curl -X POST http://localhost:9000/api/v1/admin/kb-sync
+```
+
+### 知识库同步失败
+
+```bash
+# 检查 Java 后端是否可访问
+curl http://localhost:9082/health
+
+# 检查文章接口
+curl -X POST http://localhost:9082/api/article/updated-since \
+  -H "Content-Type: application/json" \
+  -d '{"since":"2026-01-01T00:00:00","status":"published"}'
+
+# 手动触发同步
+docker-compose exec langgraph-svc python -c "
+import asyncio
+from app.tasks.knowledge_sync import sync_knowledge_base
+asyncio.run(sync_knowledge_base())
+"
+```
+
+### Java 端回退 Dify
+
+如果 Python 服务异常, Java 会自动回退 Dify:
+
+```yaml
+# application.yml 检查配置
+python:
+  enabled: true
+  circuit-breaker:
+    failure-threshold: 3
+    reset-timeout-ms: 30000
+```
+
+熔断器打开时, Java 侧日志会输出 `AiGateway 熔断器已打开`, 等待 `reset-timeout-ms` 后自动半开重试。
+
+## 扩缩容
+
+LangGraph 服务是无状态的 (ChromaDB 在共享存储上):
+
+```yaml
+# docker-compose 增加副本数
+services:
+  langgraph-svc:
+    deploy:
+      replicas: 2
+```
+
+注意: ChromaDB 文件模式不支持并发写入, 多副本时知识库同步需加锁或切 PGVector。
+
+## 备份
+
+```bash
+# ChromaDB 数据
+tar czf chroma_backup_$(date +%Y%m%d).tar.gz data/chroma_db/
+```

+ 9 - 0
cfc-langgraph/prometheus.yml

@@ -0,0 +1,9 @@
+global:
+  scrape_interval: 15s
+  evaluation_interval: 15s
+
+scrape_configs:
+  - job_name: "langgraph-svc"
+    static_configs:
+      - targets: ["langgraph-svc:9000"]
+    metrics_path: /metrics

+ 34 - 0
cfc-langgraph/pyproject.toml

@@ -0,0 +1,34 @@
+[project]
+name = "cfc-langgraph"
+version = "0.1.0"
+description = "cfc AI sidecar service with LangGraph"
+requires-python = ">=3.11"
+dependencies = [
+    "fastapi>=0.115",
+    "uvicorn[standard]>=0.34",
+    "pydantic>=2.10",
+    "pydantic-settings>=2.7",
+    "httpx>=0.28",
+    "langchain>=0.3,<1.0",
+    "langchain-community>=0.3,<1.0",
+    "langchain-openai>=0.3,<1.0",
+    "langgraph>=0.3,<1.0",
+    "chromadb>=0.6,<1.0",
+    "langchain-chroma>=0.2,<1.0",
+    "python-multipart>=0.0.20",
+    "prometheus-client>=0.21",
+]
+
+[project.optional-dependencies]
+dev = [
+    "pytest>=8",
+    "pytest-asyncio>=0.25",
+    "ruff>=0.8",
+    "httpx",
+]
+
+[tool.ruff]
+target-version = "py311"
+line-length = 100
+select = ["E", "F", "I", "N", "W"]
+ignore = ["E501"]

+ 0 - 2
cfc-langgraph/src/app.py

@@ -9,7 +9,6 @@ from fastapi.responses import JSONResponse
 
 from .schemas.questionnaire import GenerateRequest, GenerateResponse
 from .graphs.questionnaire import get_questionnaire_graph
-from .qna.router import router as qna_router
 
 router = APIRouter(prefix="/api/v1", tags=["questionnaire"])
 
@@ -43,7 +42,6 @@ async def generate_questionnaire(req: GenerateRequest):
 
 app = FastAPI(title="CFC LangGraph 问卷生成服务")
 app.include_router(router)
-app.include_router(qna_router)
 
 
 @app.get("/health")

+ 14 - 0
cfc-langgraph/start.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+# LangGraph 服务启动脚本
+# 用法: ./start.sh
+
+cd "$(dirname "$0")"
+rm -rf data/chroma_db data/chroma_db_memory
+nohup .venv/bin/uvicorn app.main:app \
+  --host 0.0.0.0 --port 9000 \
+  --workers 1 --log-level info \
+  > /tmp/langgraph.log 2>&1 &
+
+echo "PID: $!"
+sleep 3
+curl -s http://localhost:9000/api/v1/health

+ 16 - 0
cfc-langgraph/tests/conftest.py

@@ -0,0 +1,16 @@
+import pytest
+from httpx import AsyncClient, ASGITransport
+from app.main import app
+
+
+@pytest.fixture
+async def client():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as ac:
+        yield ac
+
+
+@pytest.fixture
+async def health_response(client):
+    resp = await client.get("/health")
+    return resp

+ 30 - 0
cfc-langgraph/tests/test_analyze.py

@@ -0,0 +1,30 @@
+import pytest
+from httpx import AsyncClient, ASGITransport
+from app.main import app
+
+
+@pytest.mark.asyncio
+async def test_analyze_endpoint():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/analyze", json={
+            "report_id": 1,
+            "user_id": 1,
+            "focus": "overall",
+        })
+        assert resp.status_code == 200
+        data = resp.json()
+        assert "analysis" in data
+
+
+@pytest.mark.asyncio
+async def test_analyze_missing_report():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/analyze", json={
+            "report_id": -1,
+            "user_id": 1,
+        })
+        assert resp.status_code == 200
+        data = resp.json()
+        assert isinstance(data.get("analysis"), str)

+ 151 - 0
cfc-langgraph/tests/test_auto_tag.py

@@ -0,0 +1,151 @@
+"""
+LangGraph 自动打标签 (auto-tag) 功能测试
+=============================================
+
+背景:
+    Java 后端在知识中心发布/更新文章时, 会调用 LangGraph 的 POST /api/v1/chat
+    端点, 传入"提取文章标签"的 prompt, 让 LLM 返回 3-5 个中文关键词标签。
+
+    Java 端构造的 prompt (见 ArticleService.autoGenerateTags()):
+        "你是一个家庭教育内容分析助手。请根据以下文章内容,提取 3 到 5 个
+         最能代表文章主题的中文关键词作为标签。
+         要求:1) 关键词为中文短语或名词,2-6 个字;2) 用英文逗号分隔;
+         3) 不要包含编号、引号或解释文字;4) 只返回标签列表本身。
+         文章内容:{plainText}"
+
+    Java 端解析逻辑 (与下方 parse_tags 保持一致):
+        - 用英文逗号 split
+        - 去掉行首的数字/点/横线/空白/顿号句号逗号
+        - 去掉行尾空白/顿号句号逗号
+        - 标签非空、长度 <= 20、最多取 5 个
+
+本测试: 复用 ASGI 内嵌调用, 验证 LangGraph 返回的标签质量与格式,
+        并验证其能被 Java 端的解析逻辑正确解析。
+"""
+
+import re
+
+import pytest
+from httpx import AsyncClient, ASGITransport
+
+from app.main import app
+
+
+# ============================================================
+# 测试数据
+# ============================================================
+
+# 真实文章内容 (来自 articles/D14*, 肠道菌群/过敏主题)
+ARTICLE_GUT = (
+    "儿童过敏反复不好,根源可能在肠道。人体约70%的免疫细胞分布在肠道,"
+    "肠道是人体最大的免疫器官。当肠道菌群处于健康平衡状态时,免疫系统能够准确区分"
+    "朋友和敌人。但当菌群失衡——有益菌减少、有害菌过度繁殖——肠道屏障功能就会受损。"
+    "科学调理四步走:第一步消除破坏因素,减少精制糖和加工食品;第二步重建菌群平衡,"
+    "增加膳食纤维摄入;第三步修复肠道屏障,补充谷氨酰胺、锌、维生素D;第四步维持健康"
+    "生活方式。每个孩子的菌群状况都是独特的,这正是菌群检测的核心价值。"
+)
+
+# 短内容
+ARTICLE_SHORT = "孩子睡眠质量差,作息不规律怎么办?规律作息对成长很重要。"
+
+# ============================================================
+# 工具函数
+# ============================================================
+
+# 与 Java 端 ArticleService.autoGenerateTags() 的解析逻辑保持一致
+TAG_CLEAN_START = re.compile(r"^[\d\.\-\s、。,]+")
+TAG_CLEAN_END = re.compile(r"[\s、。,]+$")
+
+
+def parse_tags(ai_response: str) -> list:
+    """按 Java 端解析规则解析 LLM 返回的标签列表"""
+    if not ai_response:
+        return []
+    tags = []
+    for raw in ai_response.split(","):
+        tag = TAG_CLEAN_START.sub("", raw.strip().strip())
+        tag = TAG_CLEAN_END.sub("", tag).strip()
+        if tag and len(tag) <= 20 and len(tags) < 5:
+            tags.append(tag)
+    return tags
+
+
+def build_tag_prompt(plain_text: str) -> str:
+    """构造与 Java 端完全一致的标签提取 prompt"""
+    return (
+        "你是一个家庭教育内容分析助手。请根据以下文章内容,"
+        "提取 3 到 5 个最能代表文章主题的中文关键词作为标签。"
+        "要求:1) 关键词为中文短语或名词,2-6 个字;2) 用英文逗号分隔;"
+        "3) 不要包含编号、引号或解释文字;4) 只返回标签列表本身。"
+        f"\n\n文章内容:{plain_text}"
+    )
+
+
+async def call_chat(query: str, user_id: int = 0) -> dict:
+    """调用 LangGraph /api/v1/chat 端点"""
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/chat", json={
+            "query": query,
+            "user_id": user_id,
+            "conversation_id": "",
+        })
+        assert resp.status_code == 200, f"chat 端点返回 {resp.status_code}"
+        return resp.json()
+
+
+# ============================================================
+# 测试用例
+# ============================================================
+
+@pytest.mark.asyncio
+async def test_auto_tag_endpoint_responds():
+    """发送标签 prompt, 端点应返回 200, 且含 answer 与 tasks 数组"""
+    data = await call_chat(build_tag_prompt(ARTICLE_GUT))
+    assert "answer" in data, "响应缺少 answer 字段"
+    assert isinstance(data.get("answer"), str)
+    assert isinstance(data.get("tasks"), list)
+
+
+@pytest.mark.asyncio
+async def test_auto_tag_valid_format():
+    """返回的标签应能被 Java 解析逻辑解析为 3-5 个中文标签"""
+    data = await call_chat(build_tag_prompt(ARTICLE_GUT))
+    tags = parse_tags(data.get("answer", ""))
+    assert 3 <= len(tags) <= 5, f"标签数量应为 3-5 个, 实际 {len(tags)}: {tags}"
+    for tag in tags:
+        # 2-6 个中文字
+        chinese_chars = re.findall(r"[\u4e00-\u9fff]", tag)
+        assert 2 <= len(chinese_chars) <= 6, f"标签应为 2-6 个中文字: '{tag}'"
+        # 不应包含编号/引号/英文逗号
+        assert not re.search(r"[\d\"',。]", tag), f"标签不应包含数字/引号/标点: '{tag}'"
+
+
+@pytest.mark.asyncio
+async def test_auto_tag_content_relevance():
+    """标签应反映文章核心主题 (肠道/过敏/免疫 等)"""
+    data = await call_chat(build_tag_prompt(ARTICLE_GUT))
+    tags = parse_tags(data.get("answer", ""))
+    assert tags, "未解析出任何标签"
+    # 核心主题词中至少命中一个
+    core_keywords = ["肠胃", "过敏", "免疫", "肠道", "菌群", "营养", "健康"]
+    hit = [kw for kw in core_keywords if any(kw in t for t in tags)]
+    assert hit, f"标签未命中核心主题词, 实际标签: {tags}"
+
+
+@pytest.mark.asyncio
+async def test_auto_tag_short_content():
+    """短内容也应正常返回标签, 不崩溃"""
+    data = await call_chat(build_tag_prompt(ARTICLE_SHORT))
+    tags = parse_tags(data.get("answer", ""))
+    assert tags, "短内容也应提取出标签"
+    assert len(tags) <= 5
+
+
+@pytest.mark.asyncio
+async def test_auto_tag_empty_content():
+    """空内容应安全返回空标签, 不崩溃"""
+    data = await call_chat(build_tag_prompt(""))
+    # 空内容不应抛异常, answer 可为空字符串
+    assert "answer" in data
+    assert isinstance(data.get("answer"), str)

+ 18 - 0
cfc-langgraph/tests/test_chat.py

@@ -0,0 +1,18 @@
+import pytest
+from httpx import AsyncClient, ASGITransport
+from app.main import app
+
+
+@pytest.mark.asyncio
+async def test_chat_endpoint():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/chat", json={
+            "query": "你好",
+            "user_id": 1,
+            "conversation_id": "",
+        })
+        assert resp.status_code == 200
+        data = resp.json()
+        assert "answer" in data
+        assert isinstance(data.get("tasks"), list)

+ 88 - 0
cfc-langgraph/tests/test_graph.py

@@ -0,0 +1,88 @@
+import json
+from unittest.mock import patch, MagicMock
+from src.schemas.questionnaire import GenerateRequest, Questionnaire, Question, Option
+
+
+def _mock_llm_response():
+    """返回合规 JSON 的 mock LLM 响应"""
+    mock_msg = MagicMock()
+    mock_msg.content = json.dumps({
+        "version": 1,
+        "questions": [
+            {
+                "id": "q1",
+                "dimension": "trust",
+                "direction": "positive",
+                "weight": 1.0,
+                "text": "您对小明的信任程度如何?",
+                "options": [
+                    {"id": "a", "score": 0},
+                    {"id": "b", "score": 1},
+                    {"id": "c", "score": 2},
+                ],
+            }
+        ]
+    }, ensure_ascii=False)
+    return mock_msg
+
+
+def _mock_llm_invalid_response():
+    mock_msg = MagicMock()
+    mock_msg.content = "这不是JSON"
+    return mock_msg
+
+
+class TestQuestionnaireSchema:
+    def test_valid_questionnaire(self):
+        q = Questionnaire(version=1, questions=[
+            Question(id="q1", dimension="trust", text="test",
+                     options=[Option(id="a", score=0)])
+        ])
+        assert q.validate_structure() is True
+
+    def test_empty_questions_fails_validation(self):
+        q = Questionnaire(version=1, questions=[])
+        assert q.validate_structure() is False
+
+    def test_question_without_options_or_scale_fails(self):
+        q = Questionnaire(version=1, questions=[
+            Question(id="q1", dimension="trust", text="test")
+        ])
+        assert q.validate_structure() is False
+
+
+class TestGraph:
+    def test_graph_valid_output(self):
+        with patch("src.graphs.questionnaire.get_llm") as mock_get_llm:
+            mock_llm = MagicMock()
+            mock_llm.invoke.return_value = _mock_llm_response()
+            mock_get_llm.return_value = mock_llm
+
+            from src.graphs.questionnaire import build_questionnaire_graph
+            graph = build_questionnaire_graph()
+            result = graph.invoke({
+                "request": GenerateRequest(member_name="小明", relationship_type="child"),
+                "raw_response": "",
+                "questionnaire": None,
+                "error": None,
+            })
+            assert result["error"] is None
+            assert result["questionnaire"] is not None
+            assert len(result["questionnaire"]["questions"]) == 1
+
+    def test_graph_invalid_llm_response_rejected(self):
+        with patch("src.graphs.questionnaire.get_llm") as mock_get_llm:
+            mock_llm = MagicMock()
+            mock_llm.invoke.return_value = _mock_llm_invalid_response()
+            mock_get_llm.return_value = mock_llm
+
+            from src.graphs.questionnaire import build_questionnaire_graph
+            graph = build_questionnaire_graph()
+            result = graph.invoke({
+                "request": GenerateRequest(member_name="小明", relationship_type="child"),
+                "raw_response": "",
+                "questionnaire": None,
+                "error": None,
+            })
+            assert result["error"] is not None
+            assert result["questionnaire"] is None

+ 37 - 0
cfc-langgraph/tests/test_intent.py

@@ -0,0 +1,37 @@
+import pytest
+from app.agents.intent_classifier import IntentClassifier, Intent
+
+
+@pytest.mark.asyncio
+async def test_classify_chat():
+    classifier = IntentClassifier()
+    intent = await classifier.classify("今天天气真好")
+    assert intent == Intent.CHAT
+
+
+@pytest.mark.asyncio
+async def test_classify_analysis():
+    classifier = IntentClassifier()
+    intent = await classifier.classify("帮我看看小明的健康报告")
+    assert intent == Intent.ANALYSIS
+
+
+@pytest.mark.asyncio
+async def test_classify_recommend():
+    classifier = IntentClassifier()
+    intent = await classifier.classify("推荐一些补钙的食物")
+    assert intent == Intent.RECOMMEND
+
+
+@pytest.mark.asyncio
+async def test_classify_mind():
+    classifier = IntentClassifier()
+    intent = await classifier.classify("孩子最近心情不好")
+    assert intent == Intent.MIND
+
+
+@pytest.mark.asyncio
+async def test_classify_health():
+    classifier = IntentClassifier()
+    intent = await classifier.classify("看看我的舌苔")
+    assert intent == Intent.HEALTH

+ 25 - 0
cfc-langgraph/tests/test_memory.py

@@ -0,0 +1,25 @@
+import pytest
+from app.memory.store import MemoryManager
+
+
+@pytest.mark.asyncio
+async def test_memory_manager_init():
+    mgr = MemoryManager()
+    assert mgr is not None
+    assert mgr.memory_vectorstore is not None
+
+
+@pytest.mark.asyncio
+async def test_working_memory():
+    mgr = MemoryManager()
+    mem = mgr.get_working_memory()
+    assert mem is not None
+    assert mem.memory_key == "history"
+
+
+@pytest.mark.asyncio
+async def test_longterm_memory():
+    mgr = MemoryManager()
+    mem = mgr.get_longterm_memory(user_id=1)
+    assert mem is not None
+    assert mem.memory_key == "long_term_memory"

+ 44 - 0
cfc-langgraph/tests/test_recommend.py

@@ -0,0 +1,44 @@
+import pytest
+from httpx import AsyncClient, ASGITransport
+from app.main import app
+
+
+@pytest.mark.asyncio
+async def test_health_check():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.get("/health")
+        assert resp.status_code == 200
+        assert resp.json() == {"status": "ok"}
+
+
+@pytest.mark.asyncio
+async def test_recommend_empty_tags():
+    """空标签应返回空列表"""
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/recommend", json={
+            "user_id": 1,
+            "tags": [],
+            "query": "推荐一些适合孩子的活动",
+            "limit": 3,
+        })
+        assert resp.status_code == 200
+        data = resp.json()
+        assert "items" in data
+
+
+@pytest.mark.asyncio
+async def test_recommend_with_tags():
+    """带标签应返回推荐结果"""
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/recommend", json={
+            "user_id": 1,
+            "tags": ["益生菌", "肠胃"],
+            "limit": 3,
+        })
+        assert resp.status_code == 200
+        data = resp.json()
+        assert "items" in data
+        assert "source" in data

+ 61 - 0
cfc-langgraph/tests/test_report_parse.py

@@ -0,0 +1,61 @@
+import pytest
+from unittest.mock import patch, AsyncMock
+from httpx import AsyncClient, ASGITransport
+from app.main import app
+
+
+@pytest.mark.asyncio
+async def test_report_parse_nonexistent_file():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/report/parse", json={
+            "file_path": "/nonexistent.pdf",
+            "family_id": 1,
+            "user_id": 1,
+        })
+        assert resp.status_code == 400
+        data = resp.json()
+        assert "文件不存在" in data["detail"]
+
+
+@pytest.mark.asyncio
+async def test_report_parse_with_real_pdf_algorithm_path():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/report/parse", json={
+            "file_path": "/app/cfc/docs/参考资料/501999942-某人.pdf",
+            "family_id": 1,
+            "user_id": 1,
+        })
+        assert resp.status_code == 200
+        data = resp.json()
+        assert data["code"] == 200
+        result = data["data"]
+        assert "format" in result
+        assert "overview" in result
+        assert "disease_risks" in result
+        assert "nutrition" in result
+        assert "amino_acids" in result
+        assert isinstance(result["overview"], dict)
+        assert isinstance(result["disease_risks"], list)
+        assert isinstance(result["amino_acids"], list)
+
+
+@pytest.mark.asyncio
+async def test_report_parse_invalid_json_response():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        with patch("app.agents.report_parse_agent.ReportParseAgent._parse_with_llm",
+                   new_callable=AsyncMock, return_value=[1, 2, 3]):
+            with patch("app.agents.report_parse_agent.parse_report_pdf_with_fallback",
+                       return_value={"format": "inline", "overview": {},
+                                     "_parse_incomplete": True}):
+                resp = await client.post("/api/v1/report/parse", json={
+                    "file_path": "/app/cfc/docs/参考资料/501999942-某人.pdf",
+                    "family_id": 1,
+                    "user_id": 1,
+                })
+                assert resp.status_code == 200
+                data = resp.json()
+                assert data["code"] == 500
+                assert "解析失败" in data["message"]

+ 12 - 0
cfc-langgraph/tests/test_tongue.py

@@ -0,0 +1,12 @@
+import pytest
+from httpx import AsyncClient, ASGITransport
+from app.main import app
+
+
+@pytest.mark.asyncio
+async def test_tongue_diagnose_missing_file():
+    """无文件上传应返回 422"""
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        resp = await client.post("/api/v1/tongue/diagnose", data={"user_id": 1})
+        assert resp.status_code == 422