From c8f26b6f9f86c39e766c49604d58a1528bdff2ce Mon Sep 17 00:00:00 2001
From: wangliang <wangliang@gaia888.com>
Date: Tue, 27 Jan 2026 13:59:12 +0800
Subject: [PATCH] =?UTF-8?q?chore:=20=E6=B8=85=E7=90=86=20scripts=20?=
 =?UTF-8?q?=E7=9B=AE=E5=BD=95=EF=BC=8C=E4=BF=9D=E7=95=99=E6=A0=B8=E5=BF=83?=
 =?UTF-8?q?=E8=BF=90=E7=BB=B4=E5=92=8C=E9=83=A8=E7=BD=B2=E8=84=9A=E6=9C=AC?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

## 变更内容

### 保留的核心运维脚本
- ✅ start.sh - 启动服务
- ✅ stop.sh - 停止服务
- ✅ init-pgvector.sql - 数据库初始化

### 保留的部署工具
- ✅ deploy-production.sh - 生产环境部署
- ✅ backup-production.sh - 生产环境备份
- ✅ set-contact-token.sh - 设置联系令牌
- ✅ set-remote-contact-token.sh - 设置远程令牌
- ✅ verify-contact-token.sh - 验证令牌

### 删除的临时调试脚本
- ❌ debug-webhook.sh - 实时监控日志
- ❌ check-conversations.sh - 检查会话
- ❌ check-chatwoot-config.sh - 检查配置
- ❌ verify-webhook.sh - 验证webhook
- ❌ update-chatwoot-webhook.sh - 更新webhook

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
---
 scripts/backup-production.sh        |  95 ++++++++++++++
 scripts/check-chatwoot-config.sh    | 108 ---------------
 scripts/check-conversations.sh      | 102 ---------------
 scripts/debug-webhook.sh            |  53 --------
 scripts/deploy-production.sh        | 195 ++++++++++++++++++++++++++++
 scripts/set-contact-token.sh        |  41 ++++++
 scripts/set-remote-contact-token.sh |  31 +++++
 scripts/start.sh                    |  73 +++++++++++
 scripts/stop.sh                     |  21 +++
 scripts/update-chatwoot-webhook.sh  |  62 ---------
 scripts/verify-contact-token.sh     |  17 +++
 scripts/verify-webhook.sh           |  81 ------------
 12 files changed, 473 insertions(+), 406 deletions(-)
 create mode 100755 scripts/backup-production.sh
 delete mode 100755 scripts/check-chatwoot-config.sh
 delete mode 100755 scripts/check-conversations.sh
 delete mode 100755 scripts/debug-webhook.sh
 create mode 100755 scripts/deploy-production.sh
 create mode 100755 scripts/set-contact-token.sh
 create mode 100755 scripts/set-remote-contact-token.sh
 create mode 100755 scripts/start.sh
 create mode 100755 scripts/stop.sh
 delete mode 100644 scripts/update-chatwoot-webhook.sh
 create mode 100755 scripts/verify-contact-token.sh
 delete mode 100755 scripts/verify-webhook.sh

diff --git a/scripts/backup-production.sh b/scripts/backup-production.sh
new file mode 100755
index 0000000..7720080
--- /dev/null
+++ b/scripts/backup-production.sh
@@ -0,0 +1,95 @@
+#!/bin/bash
+# B2B Shopping AI Assistant - Production Backup Script
+# 生产环境备份脚本
+
+set -e
+
+# 配置
+BACKUP_DIR="./backups"
+DATE=$(date +%Y%m%d_%H%M%S)
+RETENTION_DAYS=7
+
+# 颜色输出
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+NC='\033[0m'
+
+log_info() {
+    echo -e "${GREEN}[INFO]${NC} $1"
+}
+
+log_warn() {
+    echo -e "${YELLOW}[WARN]${NC} $1"
+}
+
+# 创建备份目录
+mkdir -p "$BACKUP_DIR"
+
+log_info "========================================"
+log_info "生产环境备份 - $DATE"
+log_info "========================================"
+echo ""
+
+# 1. 备份 Redis 数据
+log_info "1. 备份 Redis 数据..."
+docker run --rm \
+    -v ai_redis_data_prod:/data \
+    -v "$(pwd)/$BACKUP_DIR":/backup \
+    alpine tar czf /backup/redis-$DATE.tar.gz -C /data .
+
+log_info "✅ Redis 数据备份完成: redis-$DATE.tar.gz"
+
+# 2. 备份 Agent 日志
+log_info "2. 备份 Agent 日志..."
+docker run --rm \
+    -v ai_agent_logs_prod:/data \
+    -v "$(pwd)/$BACKUP_DIR":/backup \
+    alpine tar czf /backup/agent-logs-$DATE.tar.gz -C /data .
+
+log_info "✅ Agent 日志备份完成: agent-logs-$DATE.tar.gz"
+
+# 3. 备份 Grafana 配置（如果存在）
+if docker volume inspect ai_grafana_data &> /dev/null; then
+    log_info "3. 备份 Grafana 配置..."
+    docker run --rm \
+        -v ai_grafana_data:/data \
+        -v "$(pwd)/$BACKUP_DIR":/backup \
+        alpine tar czf /backup/grafana-$DATE.tar.gz -C /data .
+
+    log_info "✅ Grafana 配置备份完成: grafana-$DATE.tar.gz"
+fi
+
+# 4. 备份环境变量文件
+log_info "4. 备份环境变量文件..."
+cp .env.production "$BACKUP_DIR/env-$DATE.backup"
+chmod 600 "$BACKUP_DIR/env-$DATE.backup"
+
+log_info "✅ 环境变量备份完成: env-$DATE.backup"
+
+# 5. 清理旧备份
+log_info "5. 清理 $RETENTION_DAYS 天前的旧备份..."
+find "$BACKUP_DIR" -name "*.tar.gz" -mtime +$RETENTION_DAYS -delete
+find "$BACKUP_DIR" -name "env-*.backup" -mtime +$RETENTION_DAYS -delete
+
+log_info "✅ 旧备份清理完成"
+
+# 6. 生成备份清单
+log_info "6. 生成备份清单..."
+cat > "$BACKUP_DIR/manifest-$DATE.txt" << EOF
+备份时间: $DATE
+备份内容:
+- Redis 数据: redis-$DATE.tar.gz
+- Agent 日志: agent-logs-$DATE.tar.gz
+- 环境变量: env-$DATE.backup
+EOF
+
+log_info "✅ 备份清单生成完成: manifest-$DATE.txt"
+
+echo ""
+log_info "========================================"
+log_info "✅ 备份完成！"
+log_info "========================================"
+echo ""
+log_info "备份文件位置: $BACKUP_DIR"
+log_info ""
+ls -lh "$BACKUP_DIR" | grep "$DATE"
diff --git a/scripts/check-chatwoot-config.sh b/scripts/check-chatwoot-config.sh
deleted file mode 100755
index be01bc5..0000000
--- a/scripts/check-chatwoot-config.sh
+++ /dev/null
@@ -1,108 +0,0 @@
-#!/bin/bash
-# Chatwoot 配置诊断工具
-
-echo "======================================"
-echo "Chatwoot 配置诊断工具"
-echo "======================================"
-echo ""
-
-# 检查是否提供了 API Token
-if [ -z "$CHATWOOT_API_TOKEN" ]; then
-    echo "❌ 请先设置 CHATWOOT_API_TOKEN 环境变量"
-    echo ""
-    echo "获取方式："
-    echo "1. 访问 http://localhost:3000"
-    echo "2. 登录后进入 Settings → Profile → Access Tokens"
-    echo "3. 创建一个新的 Access Token"
-    echo ""
-    echo "然后运行："
-    echo "  CHATWOOT_API_TOKEN=your_token $0"
-    echo ""
-    exit 1
-fi
-
-CHATWOOT_BASE_URL="http://localhost:3000"
-ACCOUNT_ID="2"
-
-echo "🔍 正在检查 Chatwoot 配置..."
-echo ""
-
-# 1. 检查服务是否运行
-echo "1️⃣ 检查 Chatwoot 服务状态..."
-if curl -s "$CHATWOOT_BASE_URL" > /dev/null; then
-    echo "   ✅ Chatwoot 服务正常运行"
-else
-    echo "   ❌ Chatwoot 服务无法访问"
-    exit 1
-fi
-echo ""
-
-# 2. 获取所有收件箱
-echo "2️⃣ 获取所有收件箱..."
-INBOXES=$(curl -s \
-  -H "Authorization: Bearer $CHATWOOT_API_TOKEN" \
-  "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/inboxes")
-
-echo "$INBOXES" | grep -o '"id":[0-9]*' | wc -l | xargs echo "   找到收件箱数量:"
-echo ""
-
-# 3. 解析并显示每个收件箱的详细信息
-echo "3️⃣ 收件箱详细信息："
-echo "======================================"
-
-# 提取所有收件箱的 ID
-INBOX_IDS=$(echo "$INBOXES" | grep -o '"id":[0-9]*' | grep -o '[0-9]*' | sort -u)
-
-for INBOX_ID in $INBOX_IDS; do
-    echo ""
-    echo "📬 收件箱 ID: $INBOX_ID"
-    echo "--------------------------------------"
-
-    # 获取收件箱详情
-    INBOX_DETAIL=$(curl -s \
-      -H "Authorization: Bearer $CHATWOOT_API_TOKEN" \
-      "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/inboxes/$INBOX_ID")
-
-    # 提取收件箱名称
-    NAME=$(echo "$INBOX_DETAIL" | grep -o '"name":"[^"]*"' | head -1 | cut -d'"' -f4)
-    echo "   名称: $NAME"
-
-    # 提取收件箱类型
-    TYPE=$(echo "$INBOX_DETAIL" | grep -o '"inbox_type":"[^"]*"' | head -1 | cut -d'"' -f4)
-    echo "   类型: $TYPE"
-
-    # 提取 Website Token（如果有）
-    WEBSITE_TOKEN=$(echo "$INBOX_DETAIL" | grep -o '"website_token":"[^"]*"' | head -1 | cut -d'"' -f4)
-    if [ -n "$WEBSITE_TOKEN" ]; then
-        echo "   Website Token: $WEBSITE_TOKEN"
-    fi
-
-    # 提取 Webhook URL
-    WEBHOOK_URL=$(echo "$INBOX_DETAIL" | grep -o '"webhook_url":"[^"]*"' | head -1 | cut -d'"' -f4)
-    if [ -n "$WEBHOOK_URL" ]; then
-        echo "   Webhook URL: $WEBHOOK_URL"
-    else
-        echo "   Webhook URL: ❌ 未配置"
-    fi
-
-    # 检查是否是测试页面使用的 token
-    if [ "$WEBSITE_TOKEN" = "39PNCMvbMk3NvB7uaDNucc6o" ]; then
-        echo ""
-        echo "   ⭐ 这是测试页面使用的收件箱！"
-        echo "   Webhook 应该配置为: http://agent:8000/webhooks/chatwoot"
-    fi
-done
-
-echo ""
-echo "======================================"
-echo ""
-echo "📋 下一步操作："
-echo ""
-echo "1. 找到 Website Token 为 '39PNCMvbMk3NvB7uaDNucc6o' 的收件箱"
-echo "2. 记录该收件箱的 ID"
-echo "3. 确保该收件箱的 Webhook URL 配置为："
-echo "   http://agent:8000/webhooks/chatwoot"
-echo ""
-echo "💡 提示：可以通过 Chatwoot 界面更新配置："
-echo "   Settings → Inboxes → 选择收件箱 → Configuration → Webhook URL"
-echo ""
diff --git a/scripts/check-conversations.sh b/scripts/check-conversations.sh
deleted file mode 100755
index 9a4793f..0000000
--- a/scripts/check-conversations.sh
+++ /dev/null
@@ -1,102 +0,0 @@
-#!/bin/bash
-# 检查 Chatwoot 会话和消息
-
-echo "======================================"
-echo "Chatwoot 会话检查工具"
-echo "======================================"
-echo ""
-
-# 需要设置环境变量
-if [ -z "$CHATWOOT_API_TOKEN" ]; then
-    echo "❌ 请先设置 CHATWOOT_API_TOKEN 环境变量"
-    echo ""
-    echo "获取方式："
-    echo "1. 访问 http://localhost:3000"
-    echo "2. 登录后进入 Settings → Profile → Access Tokens"
-    echo "3. 创建一个新的 Access Token"
-    echo ""
-    echo "然后运行："
-    echo "  CHATWOOT_API_TOKEN=your_token $0"
-    echo ""
-    exit 1
-fi
-
-CHATWOOT_BASE_URL="http://localhost:3000"
-ACCOUNT_ID="2"
-
-echo "🔍 正在检查 Chatwoot 会话..."
-echo ""
-
-# 1. 获取所有收件箱
-echo "1️⃣ 获取所有收件箱..."
-INBOXES=$(curl -s \
-  -H "Authorization: Bearer $CHATWOOT_API_TOKEN" \
-  "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/inboxes")
-
-INBOX_IDS=$(echo "$INBOXES" | grep -o '"id":[0-9]*' | grep -o '[0-9]*' | sort -u | head -5)
-
-echo "   找到收件箱: $(echo "$INBOX_IDS" | wc -l) 个"
-echo ""
-
-# 2. 检查每个收件箱的会话
-echo "2️⃣ 检查最近的会话..."
-echo "======================================"
-
-for INBOX_ID in $INBOX_IDS; do
-    echo ""
-    echo "📬 收件箱 ID: $INBOX_ID"
-    echo "--------------------------------------"
-
-    # 获取收件箱名称
-    INBOX_NAME=$(echo "$INBOXES" | grep -o "\"id\":$INBOX_ID" -A 20 | grep '"name":"' | head -1 | cut -d'"' -f4)
-    echo "   名称: $INBOX_NAME"
-
-    # 获取最近5个会话
-    CONVERSATIONS=$(curl -s \
-      -H "Authorization: Bearer $CHATWOOT_API_TOKEN" \
-      "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/conversations?inbox_id=$INBOX_ID&sort=-created_at" | head -100)
-
-    CONV_IDS=$(echo "$CONVERSATIONS" | grep -o '"id":[0-9]*' | grep -o '[0-9]*' | head -5)
-
-    if [ -z "$CONV_IDS" ]; then
-        echo "   没有会话"
-        continue
-    fi
-
-    echo "   最近的会话:"
-    echo "$CONV_IDS" | while read CONV_ID; do
-        # 获取会话详情
-        CONV_DETAIL=$(curl -s \
-          -H "Authorization: Bearer $CHATWOOT_API_TOKEN" \
-          "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/conversations/$CONV_ID")
-
-        # 提取会话信息
-        STATUS=$(echo "$CONV_DETAIL" | grep -o '"status":"[^"]*"' | head -1 | cut -d'"' -f4)
-        CREATED_AT=$(echo "$CONV_DETAIL" | grep -o '"created_at":[^,}]*' | head -1 | cut -d'"' -f2)
-
-        # 获取消息数量
-        MESSAGES=$(curl -s \
-          -H "Authorization: Bearer $CHATWOOT_API_TOKEN" \
-          "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/conversations/$CONV_ID/messages")
-
-        MSG_COUNT=$(echo "$MESSAGES" | grep -o '"content":' | wc -l)
-
-        echo "      • 会话 #$CONV_ID - 状态: $Status - 消息数: $MSG_COUNT"
-
-        # 获取最后几条消息
-        echo "$MESSAGES" | grep -o '"content":"[^"]*"' | tail -3 | while read MSG; do
-            CONTENT=$(echo "$MSG" | cut -d'"' -f4 | sed 's/&quot;/"/g' | head -c 50)
-            echo "        - $CONTENT..."
-        done
-    done
-done
-
-echo ""
-echo "======================================"
-echo ""
-echo "💡 提示："
-echo "1. 查看上面的会话列表"
-echo "2. 记录你正在测试的会话 ID"
-echo "3. 在 Agent 日志中查找相同的 conversation_id"
-echo "4. 如果会话 ID 不匹配，说明 Widget 连接到了错误的会话"
-echo ""
diff --git a/scripts/debug-webhook.sh b/scripts/debug-webhook.sh
deleted file mode 100755
index 98ec3f7..0000000
--- a/scripts/debug-webhook.sh
+++ /dev/null
@@ -1,53 +0,0 @@
-#!/bin/bash
-# 实时监控 Chatwoot 和 Agent 日志
-
-echo "======================================"
-echo "Chatwoot 消息流程实时监控"
-echo "======================================"
-echo ""
-echo "📋 使用说明："
-echo "1. 在测试页面 http://localhost:8080/test-chat.html 发送消息"
-echo "2. 观察下面的日志输出"
-echo "3. 按 Ctrl+C 停止监控"
-echo ""
-echo "======================================"
-echo ""
-
-# 检查 Docker 容器是否运行
-if ! docker ps | grep -q "ai_agent"; then
-    echo "❌ Agent 容器未运行"
-    exit 1
-fi
-
-if ! docker ps | grep -q "ai_chatwoot"; then
-    echo "❌ Chatwoot 容器未运行"
-    exit 1
-fi
-
-echo "✅ 所有容器运行正常"
-echo ""
-echo "🔍 开始监控日志..."
-echo "======================================"
-echo ""
-
-# 使用多 tail 监控多个容器
-docker logs ai_agent -f 2>&1 &
-AGENT_PID=$!
-
-docker logs ai_chatwoot -f 2>&1 &
-CHATWOOT_PID=$!
-
-# 清理函数
-cleanup() {
-    echo ""
-    echo "======================================"
-    echo "停止监控..."
-    kill $AGENT_PID $CHATWOOT_PID 2>/dev/null
-    exit 0
-}
-
-# 捕获 Ctrl+C
-trap cleanup INT TERM
-
-# 等待
-wait
diff --git a/scripts/deploy-production.sh b/scripts/deploy-production.sh
new file mode 100755
index 0000000..a4f5b9a
--- /dev/null
+++ b/scripts/deploy-production.sh
@@ -0,0 +1,195 @@
+#!/bin/bash
+# B2B Shopping AI Assistant - Production Deployment Script
+# 生产环境部署脚本
+
+set -e  # 遇到错误立即退出
+
+# 颜色输出
+RED='\033[0;31m'
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+NC='\033[0m' # No Color
+
+# 日志函数
+log_info() {
+    echo -e "${GREEN}[INFO]${NC} $1"
+}
+
+log_warn() {
+    echo -e "${YELLOW}[WARN]${NC} $1"
+}
+
+log_error() {
+    echo -e "${RED}[ERROR]${NC} $1"
+}
+
+# 检查必要的命令
+check_requirements() {
+    log_info "检查系统依赖..."
+
+    if ! command -v docker &> /dev/null; then
+        log_error "Docker 未安装，请先安装 Docker"
+        exit 1
+    fi
+
+    if ! command -v docker-compose &> /dev/null; then
+        log_error "Docker Compose 未安装，请先安装 Docker Compose"
+        exit 1
+    fi
+
+    log_info "系统依赖检查完成"
+}
+
+# 检查环境变量文件
+check_env_file() {
+    log_info "检查环境变量文件..."
+
+    if [ ! -f .env.production ]; then
+        log_error ".env.production 文件不存在"
+        log_info "请复制 .env.production.example 并填写真实值："
+        log_info "  cp .env.production.example .env.production"
+        log_info "  vim .env.production"
+        exit 1
+    fi
+
+    # 检查必要的环境变量
+    source .env.production
+
+    required_vars=(
+        "ZHIPU_API_KEY"
+        "CHATWOOT_API_TOKEN"
+        "HYPERF_API_URL"
+        "MALL_API_URL"
+        "REDIS_PASSWORD"
+    )
+
+    missing_vars=()
+    for var in "${required_vars[@]}"; do
+        if [ -z "${!var}" ] || [[ "${!var}" == *"your_"*"_here" ]]; then
+            missing_vars+=("$var")
+        fi
+    done
+
+    if [ ${#missing_vars[@]} -ne 0 ]; then
+        log_error "以下环境变量未设置或使用默认值："
+        for var in "${missing_vars[@]}"; do
+            log_error "  - $var"
+        exit 1
+        fi
+    fi
+
+    log_info "环境变量检查完成"
+}
+
+# 构建镜像
+build_images() {
+    log_info "开始构建 Docker 镜像..."
+
+    docker-compose -f docker-compose.prod.yml build --no-cache
+
+    log_info "Docker 镜像构建完成"
+}
+
+# 停止现有服务
+stop_services() {
+    log_info "停止现有服务..."
+
+    docker-compose -f docker-compose.prod.yml down
+
+    log_info "现有服务已停止"
+}
+
+# 启动服务
+start_services() {
+    log_info "启动生产环境服务..."
+
+    docker-compose -f docker-compose.prod.yml up -d
+
+    log_info "服务启动完成"
+}
+
+# 健康检查
+health_check() {
+    log_info "等待服务启动..."
+
+    sleep 10
+
+    # 检查 Agent 服务
+    if curl -f http://localhost:8000/health &> /dev/null; then
+        log_info "✅ Agent 服务健康检查通过"
+    else
+        log_error "❌ Agent 服务健康检查失败"
+        return 1
+    fi
+
+    # 检查 MCP 服务
+    mcp_ports=(8001 8002 8003 8004)
+    for port in "${mcp_ports[@]}"; do
+        if curl -f http://localhost:$port/health &> /dev/null; then
+            log_info "✅ MCP 服务 (端口 $port) 健康检查通过"
+        else
+            log_warn "⚠️  MCP 服务 (端口 $port) 健康检查失败"
+        fi
+    done
+}
+
+# 查看服务状态
+show_status() {
+    log_info "服务状态："
+    docker-compose -f docker-compose.prod.yml ps
+}
+
+# 查看日志
+show_logs() {
+    log_info "最近的日志："
+    docker-compose -f docker-compose.prod.yml logs --tail=50
+}
+
+# 主函数
+main() {
+    log_info "========================================"
+    log_info "B2B Shopping AI Assistant - 生产环境部署"
+    log_info "========================================"
+    echo ""
+
+    # 检查参数
+    if [ "$1" == "--skip-build" ]; then
+        log_warn "跳过镜像构建步骤"
+    else
+        check_requirements
+        check_env_file
+        build_images
+    fi
+
+    stop_services
+    start_services
+
+    log_info "等待服务就绪..."
+    sleep 15
+
+    if health_check; then
+        log_info "========================================"
+        log_info "✅ 部署成功！"
+        log_info "========================================"
+        echo ""
+        log_info "服务地址："
+        log_info "  - Agent:     http://localhost:8000"
+        log_info "  - Strapi MCP: http://localhost:8001"
+        log_info "  - Order MCP:  http://localhost:8002"
+        log_info "  - After MCP:  http://localhost:8003"
+        log_info "  - Product MCP: http://localhost:8004"
+        echo ""
+        log_info "查看日志："
+        log_info "  docker-compose -f docker-compose.prod.yml logs -f"
+        echo ""
+        log_info "查看状态："
+        log_info "  docker-compose -f docker-compose.prod.yml ps"
+    else
+        log_error "部署失败，请检查日志"
+        show_logs
+        exit 1
+    fi
+}
+
+# 执行主函数
+main "$@"
diff --git a/scripts/set-contact-token.sh b/scripts/set-contact-token.sh
new file mode 100755
index 0000000..c047569
--- /dev/null
+++ b/scripts/set-contact-token.sh
@@ -0,0 +1,41 @@
+#!/bin/bash
+# 为 Chatwoot Contact 设置 JWT Token
+#
+# 使用方法:
+#   ./set-contact-token.sh <contact_id> <jwt_token>
+#
+# 示例:
+#   ./set-contact-token.sh 4 "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9..."
+
+CHATWOOT_BASE_URL="http://192.168.15.34:3000"
+ACCOUNT_ID="2"
+
+# 从环境变量或参数获取 token
+CONTACT_ID=${1:-"4"}
+JWT_TOKEN=${2:-"your_jwt_token_here"}
+MALL_TOKEN=${3:-"$JWT_TOKEN"}  # 默认使用相同的 token
+
+# Chatwoot API Token（需要在管理界面创建）
+CHATWOOT_API_TOKEN="fnWaEeAyC1gw1FYQq6YJMWSj"
+
+echo "📝 为 Contact #$CONTACT_ID 设置 token..."
+echo "JWT Token: ${JWT_TOKEN:0:30}..."
+echo "Mall Token: ${MALL_TOKEN:0:30}..."
+
+# 更新 contact 的 custom_attributes
+curl -X PUT "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/contacts/$CONTACT_ID" \
+  -H "Authorization: Bearer $CHATWOOT_API_TOKEN" \
+  -H "Content-Type: application/json" \
+  -d "{
+    \"custom_attributes\": {
+      \"jwt_token\": \"$JWT_TOKEN\",
+      \"mall_token\": \"$MALL_TOKEN\"
+    }
+  }" | python3 -m json.tool
+
+echo ""
+echo "✅ Token 设置完成！"
+echo ""
+echo "验证："
+echo "  curl -H \"Authorization: Bearer $CHATWOOT_API_TOKEN\" \\"
+echo "    $CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/contacts/$CONTACT_ID"
diff --git a/scripts/set-remote-contact-token.sh b/scripts/set-remote-contact-token.sh
new file mode 100755
index 0000000..ee6c982
--- /dev/null
+++ b/scripts/set-remote-contact-token.sh
@@ -0,0 +1,31 @@
+#!/bin/bash
+# 为远程 Chatwoot 的 Contact 设置 JWT Token
+
+CHATWOOT_BASE_URL="http://192.168.15.28:3000"
+ACCOUNT_ID="2"
+CONTACT_ID=${1:-"4"}  # Contact ID (从日志中看到是 4)
+
+# 从参数或使用测试 token
+JWT_TOKEN=${2:-"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzdWIiOiIxMjM0NTY3ODkwIiwibmFtZSI6IkpvaG4gRG9lIiwiaWF0IjoxNTE2MjM5MDIyfQ.SflKxwRJSMeKKF2QT4fwpMeJf36POk6yJV_adQssw5c"}
+
+echo "📝 为 Contact #$CONTACT_ID 设置 token..."
+echo "JWT Token: ${JWT_TOKEN:0:30}..."
+echo ""
+
+# 更新 contact 的 custom_attributes
+curl -X PUT "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/contacts/$CONTACT_ID" \
+  -H "Authorization: Bearer wFc2Dpi3wcf9eT5Cibckd68z" \
+  -H "Content-Type: application/json" \
+  -d "{
+    \"custom_attributes\": {
+      \"jwt_token\": \"$JWT_TOKEN\",
+      \"mall_token\": \"$JWT_TOKEN\"
+    }
+  }" | python3 -m json.tool
+
+echo ""
+echo "✅ Token 设置完成！"
+echo ""
+echo "验证："
+echo "  curl -H \"Authorization: Bearer wFc2Dpi3wcf9eT5Cibckd68z\" \\"
+echo "    $CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/contacts/$CONTACT_ID"
diff --git a/scripts/start.sh b/scripts/start.sh
new file mode 100755
index 0000000..b09d56b
--- /dev/null
+++ b/scripts/start.sh
@@ -0,0 +1,73 @@
+#!/bin/bash
+# 启动脚本 - B2B AI Assistant
+
+set -e
+
+echo "========================================"
+echo "🚀 启动 B2B AI Assistant 服务"
+echo "========================================"
+echo ""
+
+# 检查 Docker 是否运行
+if ! docker info > /dev/null 2>&1; then
+    echo "❌ Docker 未运行，请先启动 Docker"
+    exit 1
+fi
+
+echo "✅ Docker 运行正常"
+echo ""
+
+# 检查远程 Chatwoot 连接
+echo "🔍 检查远程 Chatwoot 连接..."
+CHATWOOT_URL="http://192.168.15.28:3000"
+if curl -s --connect-timeout 5 "$CHATWOOT_URL" > /dev/null 2>&1; then
+    echo "✅ 远程 Chatwoot 连接正常 ($CHATWOOT_URL)"
+else
+    echo "⚠️  警告：无法连接到远程 Chatwoot ($CHATWOOT_URL)"
+    echo "   请确保 Chatwoot 正在运行"
+    read -p "是否继续启动？(y/n) " -n 1 -r
+    echo
+    if [[ ! $REPLY =~ ^[Yy]$ ]]; then
+        exit 1
+    fi
+fi
+echo ""
+
+# 获取本机 IP
+LOCAL_IP=$(hostname -I | awk '{print $1}')
+echo "📡 本机 IP: $LOCAL_IP"
+echo "   Webhook URL: http://$LOCAL_IP:8000/webhooks/chatwoot"
+echo ""
+
+# 启动服务
+echo "🐳 启动 Docker 服务..."
+docker-compose up -d
+
+echo ""
+echo "⏳ 等待服务启动..."
+sleep 5
+
+# 检查服务状态
+echo ""
+echo "📊 服务状态："
+docker-compose ps
+
+echo ""
+echo "========================================"
+echo "✅ 服务启动完成！"
+echo "========================================"
+echo ""
+echo "📝 常用命令："
+echo "  查看日志: docker-compose logs -f agent"
+echo "  查看状态: docker-compose ps"
+echo "  停止服务: docker-compose down"
+echo "  重启服务: docker-compose restart [service_name]"
+echo ""
+echo "📖 测试页面："
+echo "  http://localhost:8080/docs/test-chat.html"
+echo "  http://localhost:8080/docs/test-conversation-id.html"
+echo ""
+echo "⚠️  重要："
+echo "  请确保远程 Chatwoot (192.168.15.28:3000) 已配置 Webhook:"
+echo "  http://$LOCAL_IP:8000/webhooks/chatwoot"
+echo ""
diff --git a/scripts/stop.sh b/scripts/stop.sh
new file mode 100755
index 0000000..8bc31d1
--- /dev/null
+++ b/scripts/stop.sh
@@ -0,0 +1,21 @@
+#!/bin/bash
+# 停止脚本 - B2B AI Assistant
+
+set -e
+
+echo "========================================"
+echo "🛑 停止 B2B AI Assistant 服务"
+echo "========================================"
+echo ""
+
+# 停止服务
+echo "🐳 停止 Docker 服务..."
+docker-compose down
+
+echo ""
+echo "✅ 所有服务已停止"
+echo ""
+echo "💡 提示："
+echo "  如需删除数据卷，运行: docker-compose down -v"
+echo "  如需重新启动，运行: ./scripts/start.sh"
+echo ""
diff --git a/scripts/update-chatwoot-webhook.sh b/scripts/update-chatwoot-webhook.sh
deleted file mode 100644
index d1a3ddb..0000000
--- a/scripts/update-chatwoot-webhook.sh
+++ /dev/null
@@ -1,62 +0,0 @@
-#!/bin/bash
-# 更新 Chatwoot Webhook 配置脚本
-
-# 配置
-CHATWOOT_BASE_URL="http://localhost:3000"
-ACCOUNT_ID="2"  # 你的账户 ID
-INBOX_ID=""     # 需要填入你的收件箱 ID
-API_TOKEN=""    # 需要填入你的 Chatwoot API Token
-NEW_WEBHOOK_URL="http://agent:8000/webhooks/chatwoot"
-WEBHOOK_SECRET="b7a12b9c9173718596f02fd912fb59f97891a0e7abb1a5e457b4c8858b2d21b5"
-
-# 使用说明
-echo "======================================"
-echo "Chatwoot Webhook 配置更新工具"
-echo "======================================"
-echo ""
-echo "请先设置以下变量："
-echo "1. INBOX_ID - 你的收件箱 ID"
-echo "2. API_TOKEN - Chatwoot API Token（从 Settings → Profile → Access Tokens 获取）"
-echo ""
-echo "然后运行："
-echo "  INBOX_ID=<收件箱ID> API_TOKEN=<API Token> $0"
-echo ""
-echo "或者直接编辑此脚本设置变量。"
-echo ""
-
-# 检查参数
-if [ -z "$INBOX_ID" ] || [ -z "$API_TOKEN" ]; then
-    echo "❌ 缺少必要参数"
-    exit 1
-fi
-
-# 获取当前 webhook 配置
-echo "📋 获取当前 webhook 配置..."
-CURRENT_CONFIG=$(curl -s \
-  -H "Authorization: Bearer $API_TOKEN" \
-  "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/inboxes/$INBOX_ID")
-
-echo "当前配置："
-echo "$CURRENT_CONFIG" | grep -o '"webhook_url":"[^"]*"' || echo "未找到 webhook_url"
-
-# 更新 webhook
-echo ""
-echo "🔄 更新 webhook URL 为: $NEW_WEBHOOK_URL"
-
-UPDATE_RESPONSE=$(curl -s -X PUT \
-  -H "Authorization: Bearer $API_TOKEN" \
-  -H "Content-Type: application/json" \
-  -d "{
-    \"inbox\": {
-      \"webhook_url\": \"$NEW_WEBHOOK_URL\"
-    }
-  }" \
-  "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/inboxes/$INBOX_ID")
-
-echo "更新响应："
-echo "$UPDATE_RESPONSE"
-
-echo ""
-echo "✅ 配置更新完成！"
-echo ""
-echo "现在可以在 Chatwoot 中测试发送消息了。"
diff --git a/scripts/verify-contact-token.sh b/scripts/verify-contact-token.sh
new file mode 100755
index 0000000..07de1f0
--- /dev/null
+++ b/scripts/verify-contact-token.sh
@@ -0,0 +1,17 @@
+#!/bin/bash
+# 验证 Contact 的 token 设置
+
+CHATWOOT_BASE_URL="http://192.168.15.34:3000"
+ACCOUNT_ID="2"
+CONTACT_ID=${1:-"4"}
+CHATWOOT_API_TOKEN="fnWaEeAyC1gw1FYQq6YJMWSj"
+
+echo "🔍 查询 Contact #$CONTACT_ID 的信息..."
+echo ""
+
+curl -s "$CHATWOOT_BASE_URL/api/v1/accounts/$ACCOUNT_ID/contacts/$CONTACT_ID" \
+  -H "Authorization: Bearer $CHATWOOT_API_TOKEN" \
+  -H "Content-Type: application/json" | python3 -m json.tool | grep -A 20 "custom_attributes"
+
+echo ""
+echo "✅ 查询完成！"
diff --git a/scripts/verify-webhook.sh b/scripts/verify-webhook.sh
deleted file mode 100755
index b309059..0000000
--- a/scripts/verify-webhook.sh
+++ /dev/null
@@ -1,81 +0,0 @@
-#!/bin/bash
-# 验证 Chatwoot Webhook 配置
-
-echo "======================================"
-echo "Chatwoot Webhook 配置验证工具"
-echo "======================================"
-echo ""
-
-# 检查 Agent 服务
-echo "1️⃣ 检查 Agent 服务..."
-if curl -s http://localhost:8000/health | grep -q "healthy"; then
-    echo "   ✅ Agent 服务运行正常 (http://localhost:8000)"
-else
-    echo "   ❌ Agent 服务未运行"
-    exit 1
-fi
-echo ""
-
-# 检查 Chatwoot 服务
-echo "2️⃣ 检查 Chatwoot 服务..."
-if curl -s http://localhost:3000 > /dev/null; then
-    echo "   ✅ Chatwoot 服务运行正常 (http://localhost:3000)"
-else
-    echo "   ❌ Chatwoot 服务未运行"
-    exit 1
-fi
-echo ""
-
-# 检查网络连通性（从 Chatwoot 容器访问 Agent）
-echo "3️⃣ 检查容器间网络连通性..."
-if docker exec ai_chatwoot wget -q -O - http://agent:8000/health | grep -q "healthy"; then
-    echo "   ✅ Chatwoot 可以访问 Agent (http://agent:8000)"
-else
-    echo "   ❌ Chatwoot 无法访问 Agent"
-    echo "   请检查两个容器是否在同一 Docker 网络中"
-    exit 1
-fi
-echo ""
-
-# 检查环境变量配置
-echo "4️⃣ 检查环境变量配置..."
-if [ -f .env ]; then
-    if grep -q "CHATWOOT_WEBHOOK_SECRET" .env; then
-        echo "   ✅ CHATWOOT_WEBHOOK_SECRET 已配置"
-    else
-        echo "   ⚠️  CHATWOOT_WEBHOOK_SECRET 未配置（可选）"
-    fi
-else
-    echo "   ⚠️  .env 文件不存在"
-fi
-echo ""
-
-# 显示配置摘要
-echo "======================================"
-echo "📋 配置摘要"
-echo "======================================"
-echo ""
-echo "Agent 服务:"
-echo "  • 容器名称: ai_agent"
-echo "  • 内部地址: http://agent:8000"
-echo "  • Webhook 端点: http://agent:8000/webhooks/chatwoot"
-echo "  • 外部访问: http://localhost:8000"
-echo ""
-echo "Chatwoot 服务:"
-echo "  • 容器名称: ai_chatwoot"
-echo "  • 内部地址: http://chatwoot:3000"
-echo "  • 外部访问: http://localhost:3000"
-echo ""
-echo "📝 在 Chatwoot 界面中配置:"
-echo "  1. 访问: http://localhost:3000"
-echo "  2. 进入: Settings → Inboxes → 选择 Website 收件箱"
-echo "  3. 点击: Configuration 标签"
-echo "  4. 设置 Webhook URL 为: http://agent:8000/webhooks/chatwoot"
-echo "  5. 点击 Save 保存"
-echo ""
-echo "⚠️  注意事项:"
-echo "  • 不要在 Chatwoot 中启用内置机器人（Bot）"
-echo "  • 只配置 Webhook 即可"
-echo "  • Webhook URL 使用 'agent' 而不是 'localhost'"
-echo ""
-echo "======================================"