diff --git a/README.md b/README.md
index 4f8916b..be428ca 100644
--- a/README.md
+++ b/README.md
@@ -19,16 +19,58 @@
 
 ## 🚀 快速开始
 
-### 环境要求
+### 方法 1: 自动化脚本（推荐） ⚡
 
+**一键完成所有设置、启动并测试服务器：**
+
+```bash
+# 1. 克隆项目
+git clone https://github.com/Zeeeepa/z.ai2api_python.git
+cd z.ai2api_python
+
+# 2. 设置环境变量
+export ZAI_EMAIL=your-email@example.com
+export ZAI_PASSWORD=your-password
+
+# 3. 运行自动化脚本
+bash scripts/all.sh
+```
+
+这将自动完成：
+- ✅ 安装所有依赖（Python + Playwright）
+- ✅ 通过 Playwright 自动登录并获取认证令牌
+- ✅ 在 8080 端口启动服务器
+- ✅ 发送测试请求验证 API 功能
+- ✅ 显示实时日志
+
+**单独运行脚本：**
+
+```bash
+# 仅环境设置和令牌获取
+bash scripts/setup.sh
+
+# 仅启动服务器
+bash scripts/start.sh
+
+# 仅测试 API
+bash scripts/send_openai_request.sh
+```
+
+📖 详细文档请参阅 [scripts/README.md](scripts/README.md)
+
+---
+
+### 方法 2: 手动安装
+
+**环境要求：**
 - Python 3.9-3.12
 - pip 或 uv (推荐)
 
-### 本地运行
+**步骤：**
 
 ```bash
 # 1. 克隆项目
-git clone https://github.com/ZyphrZero/z.ai2api_python.git
+git clone https://github.com/Zeeeepa/z.ai2api_python.git
 cd z.ai2api_python
 
 # 2. 安装依赖（使用 uv 推荐）
diff --git a/scripts/README.md b/scripts/README.md
new file mode 100644
index 0000000..b5c1629
--- /dev/null
+++ b/scripts/README.md
@@ -0,0 +1,439 @@
+# Z.AI2API Automation Scripts
+
+Complete automation suite for setting up, starting, and testing the Z.AI2API Python server.
+
+## 🔥 IMPORTANT: NO MOCKS - 100% REAL API CALLS
+
+**All scripts use ACTUAL API calls with REAL responses from Z.AI!**
+
+The `send_openai_request.sh` script makes genuine requests to your running server, which:
+- ✅ **Connects to real Z.AI chat interface** (not a mock)
+- ✅ **Uses your JWT token** from browser session
+- ✅ **Mimics web browser behavior** exactly
+- ✅ **Returns authentic AI responses** from Z.AI/GLM models
+- ✅ **Converts responses** to OpenAI API format
+
+**You see the actual AI-generated content, not fake data!**
+
+## 📋 Overview
+
+This directory contains scripts to automate the entire workflow:
+
+1. **setup.sh** - Environment setup and token retrieval
+2. **start.sh** - Server startup and health monitoring
+3. **send_openai_request.sh** - API testing with **REAL** OpenAI-compatible requests
+4. **all.sh** - Complete orchestration of all steps
+5. **retrieve_token.py** - Playwright-based token retrieval
+
+## 🚀 Quick Start
+
+### Prerequisites
+
+- Python 3.9-3.12
+- Bash shell
+- curl
+- Internet connection
+
+### Environment Variables
+
+Set these before running the scripts:
+
+```bash
+export ZAI_EMAIL=your-email@example.com
+export ZAI_PASSWORD=your-password
+```
+
+### One-Command Setup
+
+Run everything at once:
+
+```bash
+git clone https://github.com/Zeeeepa/z.ai2api_python.git
+cd z.ai2api_python
+export ZAI_EMAIL=your-email@example.com
+export ZAI_PASSWORD=your-password
+bash scripts/all.sh
+```
+
+This will:
+- ✅ Install all dependencies
+- ✅ Retrieve authentication token automatically via Playwright
+- ✅ Start the server on port 8080
+- ✅ Test the API with **REAL requests to Z.AI** (no mocks!)
+- ✅ Display **actual AI-generated responses**
+- ✅ Show live logs
+
+## 📖 Individual Scripts
+
+### 1. setup.sh
+
+Performs initial environment setup:
+
+```bash
+export ZAI_EMAIL=your-email@example.com
+export ZAI_PASSWORD=your-password
+bash scripts/setup.sh
+```
+
+**What it does:**
+- Checks Python version (requires 3.9+)
+- Installs Python dependencies (via pip or uv)
+- Installs Playwright and Chromium browser
+- Creates `.env` file from template
+- Retrieves authentication token via automated browser login
+- Saves token to `.env` file
+- Initializes SQLite database
+
+**Requirements:**
+- `ZAI_EMAIL` environment variable
+- `ZAI_PASSWORD` environment variable
+
+**Output:**
+- `.env` file with valid `AUTH_TOKEN`
+- `tokens.db` database file
+- Installed dependencies
+
+### 2. start.sh
+
+Starts the API server:
+
+```bash
+bash scripts/start.sh
+```
+
+**What it does:**
+- Checks if `.env` exists
+- Validates port availability (default: 8080)
+- Verifies dependencies are installed
+- Starts server in background
+- Monitors server health
+- Creates PID file for process management
+- Shows live logs
+
+**Requirements:**
+- `.env` file with valid configuration
+- Available port (check `LISTEN_PORT` in `.env`)
+
+**Output:**
+- Running server on http://localhost:8080
+- PID file: `.server.pid`
+- Log file: `logs/server.log`
+
+**Useful commands after starting:**
+```bash
+# View logs
+tail -f logs/server.log
+
+# Stop server
+kill $(cat .server.pid)
+
+# Check if running
+curl http://localhost:8080/
+```
+
+### 3. send_openai_request.sh
+
+Tests the API with **REAL** OpenAI-compatible requests:
+
+```bash
+bash scripts/send_openai_request.sh
+```
+
+**What it does:**
+- Checks if server is running
+- Sends **ACTUAL API calls** to your running server
+- Server proxies to **real Z.AI chat interface** using your JWT token
+- Displays **genuine AI-generated responses** from Z.AI/GLM models
+- Tests streaming mode with **real streaming responses**
+- Validates complete API functionality
+
+**NO MOCKS:** All responses are authentic AI-generated content from Z.AI's service!
+
+**Requirements:**
+- Server running on configured port
+- Valid `AUTH_TOKEN` in `.env`
+
+**Output:**
+- Formatted API response
+- Token usage statistics
+- Streaming test results
+
+### 4. all.sh
+
+Orchestrates complete workflow:
+
+```bash
+export ZAI_EMAIL=your-email@example.com
+export ZAI_PASSWORD=your-password
+bash scripts/all.sh
+```
+
+**What it does:**
+- Runs `setup.sh` to configure environment
+- Runs `start.sh` to start server
+- Runs `send_openai_request.sh` to test API
+- Displays comprehensive status
+- Shows live logs
+
+**Requirements:**
+- Same as `setup.sh`
+
+**Output:**
+- Running and tested server
+- Live log display
+
+### 5. retrieve_token.py
+
+Python script for automated token retrieval (called by setup.sh):
+
+```bash
+python scripts/retrieve_token.py
+```
+
+**What it does:**
+- Launches headless Chromium browser
+- Navigates to Z.AI login page
+- Enters credentials
+- Submits login form
+- Extracts authentication token
+- Saves token to `.env` file
+
+**Requirements:**
+- `ZAI_EMAIL` environment variable
+- `ZAI_PASSWORD` environment variable
+- Playwright installed with browsers
+
+## 🔧 Configuration
+
+### Port Configuration
+
+Edit `.env` to change the server port:
+
+```env
+LISTEN_PORT=8080
+```
+
+### Authentication
+
+The `AUTH_TOKEN` is automatically retrieved by `setup.sh`. To manually set it:
+
+```env
+AUTH_TOKEN=your-token-here
+```
+
+### Debug Mode
+
+Enable detailed logging:
+
+```env
+DEBUG_LOGGING=true
+```
+
+## 📊 Server Management
+
+### Check Server Status
+
+```bash
+# Using curl
+curl http://localhost:8080/
+
+# Check process
+ps aux | grep main.py
+```
+
+### View Logs
+
+```bash
+# Real-time logs
+tail -f logs/server.log
+
+# Recent logs
+tail -100 logs/server.log
+
+# Search logs
+grep "ERROR" logs/server.log
+```
+
+### Stop Server
+
+```bash
+# Using PID file
+kill $(cat .server.pid)
+
+# Or manually
+kill <PID>
+
+# Force kill if needed
+kill -9 <PID>
+```
+
+### Restart Server
+
+```bash
+# Stop and start
+kill $(cat .server.pid) && bash scripts/start.sh
+
+# Or just run start.sh (it will detect if already running)
+bash scripts/start.sh
+```
+
+## 🧪 Testing
+
+### Manual API Test
+
+```bash
+# Using curl
+curl -X POST http://localhost:8080/v1/chat/completions \
+  -H "Content-Type: application/json" \
+  -H "Authorization: Bearer YOUR_AUTH_TOKEN" \
+  -d '{
+    "model": "GLM-4-6-API-V1",
+    "messages": [
+      {"role": "user", "content": "Hello!"}
+    ],
+    "stream": false
+  }'
+```
+
+### Automated Test
+
+```bash
+bash scripts/send_openai_request.sh
+```
+
+## 🐛 Troubleshooting
+
+### Setup Issues
+
+**Problem:** `ZAI_EMAIL and ZAI_PASSWORD must be set`
+```bash
+# Solution: Export environment variables
+export ZAI_EMAIL=your-email@example.com
+export ZAI_PASSWORD=your-password
+```
+
+**Problem:** `Playwright browsers not found`
+```bash
+# Solution: Install browsers manually
+python3 -m playwright install chromium
+```
+
+### Server Issues
+
+**Problem:** `Port already in use`
+```bash
+# Solution 1: Kill process using the port
+lsof -ti:8080 | xargs kill
+
+# Solution 2: Change port in .env
+echo "LISTEN_PORT=8081" >> .env
+```
+
+**Problem:** `Server failed to start`
+```bash
+# Check logs
+cat logs/server.log
+
+# Check dependencies
+python3 -c "import fastapi, granian, httpx"
+```
+
+### Token Issues
+
+**Problem:** `Failed to retrieve token`
+```bash
+# Solution 1: Check credentials
+echo $ZAI_EMAIL
+echo $ZAI_PASSWORD
+
+# Solution 2: Run token retrieval with verbose output
+python3 scripts/retrieve_token.py
+
+# Solution 3: Manually set token in .env
+echo "AUTH_TOKEN=your-token" >> .env
+```
+
+### API Test Issues
+
+**Problem:** `Authentication failed`
+```bash
+# Check AUTH_TOKEN in .env
+grep AUTH_TOKEN .env
+
+# Test with different token
+export AUTH_TOKEN=your-token
+bash scripts/send_openai_request.sh
+```
+
+## 📝 Script Details
+
+### Directory Structure
+
+```
+scripts/
+├── README.md                    # This file
+├── all.sh                       # Complete orchestration
+├── setup.sh                     # Environment setup
+├── start.sh                     # Server startup
+├── send_openai_request.sh      # API testing
+└── retrieve_token.py           # Token retrieval
+```
+
+### Script Dependencies
+
+```
+all.sh
+├── setup.sh
+│   ├── retrieve_token.py
+│   └── Python dependencies
+├── start.sh
+│   └── main.py
+└── send_openai_request.sh
+    └── curl
+```
+
+## 🔐 Security Notes
+
+- **Never commit** `.env` file to version control
+- **Keep** `AUTH_TOKEN` private
+- **Use** strong passwords for ZAI_EMAIL account
+- **Rotate** tokens regularly
+- **Monitor** `logs/server.log` for suspicious activity
+
+## 🎯 Best Practices
+
+1. **Always use environment variables** for credentials
+2. **Check logs** regularly: `tail -f logs/server.log`
+3. **Monitor server health** periodically
+4. **Keep dependencies updated**: `pip install -U -r requirements.txt`
+5. **Use setup.sh** for fresh installations
+6. **Use all.sh** for quick testing
+
+## 📚 Additional Resources
+
+- **Project README**: `../README.md`
+- **API Documentation**: http://localhost:8080/docs
+- **Admin Panel**: http://localhost:8080/admin
+- **GitHub**: https://github.com/Zeeeepa/z.ai2api_python
+
+## 💡 Tips
+
+- Use `uv` for faster dependency management
+- Enable `DEBUG_LOGGING=true` when troubleshooting
+- Check `logs/server.log` for detailed error messages
+- Use `send_openai_request.sh` to verify changes
+- Keep `ZAI_PASSWORD` secure
+
+## 🆘 Getting Help
+
+If you encounter issues:
+
+1. Check the troubleshooting section above
+2. Review `logs/server.log` for errors
+3. Ensure all prerequisites are met
+4. Verify environment variables are set correctly
+5. Open an issue on GitHub
+
+---
+
+**Happy coding! 🚀**
diff --git a/scripts/all.sh b/scripts/all.sh
new file mode 100755
index 0000000..f50904c
--- /dev/null
+++ b/scripts/all.sh
@@ -0,0 +1,280 @@
+#!/bin/bash
+#
+# all.sh - Complete Z.AI2API Setup, Start, and Test Script
+#
+# This script orchestrates the complete workflow with REAL API calls:
+# 
+# 1. Run setup.sh
+#    ✅ Install dependencies
+#    ✅ Use Playwright to login and retrieve AUTH_TOKEN
+#    ✅ Save token to .env file
+#
+# 2. Run start.sh
+#    ✅ Start server using the retrieved AUTH_TOKEN
+#    ✅ Server proxies to real Z.AI chat interface
+#
+# 3. Run send_openai_request.sh
+#    ✅ Send REAL OpenAI API request: "What is Graph-RAG?"
+#    ✅ Display actual AI-generated response from Z.AI
+#
+# Usage:
+#   export ZAI_EMAIL=your-email@example.com
+#   export ZAI_PASSWORD=your-password
+#   bash scripts/all.sh
+#
+
+set -e  # Exit on error
+
+# Colors for output
+RED='\033[0;31m'
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+BLUE='\033[0;34m'
+MAGENTA='\033[0;35m'
+CYAN='\033[0;36m'
+NC='\033[0m' # No Color
+
+# Script directory
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+PROJECT_ROOT="$(dirname "$SCRIPT_DIR")"
+
+# Logging functions
+log_info() {
+    echo -e "${BLUE}ℹ️  $1${NC}"
+}
+
+log_success() {
+    echo -e "${GREEN}✅ $1${NC}"
+}
+
+log_warning() {
+    echo -e "${YELLOW}⚠️  $1${NC}"
+}
+
+log_error() {
+    echo -e "${RED}❌ $1${NC}"
+}
+
+log_step() {
+    echo -e "${MAGENTA}$1${NC}"
+}
+
+# Error handler
+cleanup_on_error() {
+    echo ""
+    log_error "An error occurred during execution"
+    
+    # Stop server if it was started
+    if [ -f "$PROJECT_ROOT/.server.pid" ]; then
+        PID=$(cat "$PROJECT_ROOT/.server.pid")
+        if kill -0 $PID 2>/dev/null; then
+            log_info "Stopping server (PID: $PID)..."
+            kill $PID 2>/dev/null || true
+            rm -f "$PROJECT_ROOT/.server.pid"
+        fi
+    fi
+    
+    echo ""
+    log_warning "Please check the error messages above and try again"
+    exit 1
+}
+
+trap cleanup_on_error ERR
+
+# Banner
+clear
+echo ""
+echo "╔════════════════════════════════════════════════════════════╗"
+echo "║                                                            ║"
+echo "║         Z.AI2API Python - Complete Setup & Test           ║"
+echo "║                                                            ║"
+echo "╚════════════════════════════════════════════════════════════╝"
+echo ""
+
+# Pre-flight checks
+log_info "Performing pre-flight checks..."
+
+# Check if required environment variables are set
+if [ -z "$ZAI_EMAIL" ] || [ -z "$ZAI_PASSWORD" ]; then
+    log_error "Required environment variables not set"
+    echo ""
+    echo "Please export the following environment variables:"
+    echo ""
+    echo "  export ZAI_EMAIL=your-email@example.com"
+    echo "  export ZAI_PASSWORD=your-password"
+    echo ""
+    echo "Then run:"
+    echo "  bash scripts/all.sh"
+    echo ""
+    exit 1
+fi
+
+log_success "Environment variables verified"
+log_info "Email: $ZAI_EMAIL"
+echo ""
+
+# Confirm with user
+echo "This script will:"
+echo "  1️⃣  Install dependencies and retrieve authentication token"
+echo "  2️⃣  Start the API server using the retrieved token"
+echo "  3️⃣  Send REAL API request: 'What is Graph-RAG?'"
+echo "  4️⃣  Display actual AI response from Z.AI"
+echo ""
+read -p "Continue? (y/N): " -n 1 -r
+echo ""
+
+if [[ ! $REPLY =~ ^[Yy]$ ]]; then
+    log_warning "Aborted by user"
+    exit 0
+fi
+
+echo ""
+echo "============================================================"
+log_step "STEP 1/3: Running setup.sh"
+echo "============================================================"
+echo ""
+
+# Execute setup script (this will create and activate venv)
+bash "$SCRIPT_DIR/setup.sh"
+
+if [ $? -ne 0 ]; then
+    log_error "Setup failed"
+    exit 1
+fi
+
+# Ensure venv is activated for subsequent steps
+VENV_DIR="$PROJECT_ROOT/.venv"
+if [ -d "$VENV_DIR" ]; then
+    log_info "Activating virtual environment for subsequent steps..."
+    source "$VENV_DIR/bin/activate"
+    log_success "Virtual environment activated"
+fi
+
+echo ""
+echo "============================================================"
+log_step "STEP 2/3: Running start.sh"
+echo "============================================================"
+echo ""
+
+# Execute start script
+# Note: We need to handle this specially since start.sh might tail logs
+# We'll start it in a way that returns control to us
+
+# Kill any tail processes from previous runs
+pkill -f "tail -f logs/server.log" 2>/dev/null || true
+
+# Start server without tailing logs
+bash "$SCRIPT_DIR/start.sh" > /tmp/start_output.log 2>&1 &
+START_PID=$!
+
+# Wait for startup to complete
+sleep 5
+
+# Check if start script succeeded
+if ! wait $START_PID 2>/dev/null; then
+    # Check if server is actually running
+    if [ -f "$PROJECT_ROOT/.server.pid" ]; then
+        SERVER_PID=$(cat "$PROJECT_ROOT/.server.pid")
+        if kill -0 $SERVER_PID 2>/dev/null; then
+            log_success "Server started successfully (PID: $SERVER_PID)"
+        else
+            log_error "Server failed to start"
+            cat /tmp/start_output.log
+            exit 1
+        fi
+    else
+        log_error "Server startup failed"
+        cat /tmp/start_output.log
+        exit 1
+    fi
+fi
+
+# Additional health check
+log_info "Verifying server health..."
+MAX_RETRIES=10
+RETRY=0
+PORT=${LISTEN_PORT:-8080}
+
+while [ $RETRY -lt $MAX_RETRIES ]; do
+    if curl -s -f "http://localhost:$PORT/" > /dev/null 2>&1; then
+        log_success "Server is healthy and responding"
+        break
+    fi
+    RETRY=$((RETRY + 1))
+    echo -n "."
+    sleep 1
+done
+
+if [ $RETRY -eq $MAX_RETRIES ]; then
+    log_error "Server health check failed"
+    exit 1
+fi
+
+echo ""
+sleep 2  # Give server a moment to fully initialize
+
+echo ""
+echo "============================================================"
+log_step "STEP 3/3: Running send_openai_request.sh"
+echo "============================================================"
+echo ""
+
+# Execute test request script
+bash "$SCRIPT_DIR/send_openai_request.sh"
+
+if [ $? -ne 0 ]; then
+    log_error "API test failed"
+    exit 1
+fi
+
+# Final success message
+echo ""
+echo "╔════════════════════════════════════════════════════════════╗"
+echo "║                                                            ║"
+echo "║           🎉 All Steps Completed Successfully! 🎉          ║"
+echo "║                                                            ║"
+echo "╚════════════════════════════════════════════════════════════╝"
+echo ""
+echo -e "${GREEN}✨ Your Z.AI2API server is now running and verified! ✨${NC}"
+echo ""
+echo "📊 Server Status:"
+if [ -f "$PROJECT_ROOT/.server.pid" ]; then
+    SERVER_PID=$(cat "$PROJECT_ROOT/.server.pid")
+    echo "  Status:     Running"
+    echo "  PID:        $SERVER_PID"
+else
+    echo "  Status:     Unknown (PID file not found)"
+fi
+echo "  Port:       ${PORT:-8080}"
+echo "  Log:        logs/server.log"
+echo ""
+echo "🌐 Access Points:"
+echo "  API Root:   http://localhost:${PORT:-8080}/"
+echo "  API Docs:   http://localhost:${PORT:-8080}/docs"
+echo "  Admin:      http://localhost:${PORT:-8080}/admin"
+echo ""
+echo "📝 Useful Commands:"
+echo "  View logs:          tail -f logs/server.log"
+echo "  Test API again:     bash scripts/send_openai_request.sh"
+if [ -f "$PROJECT_ROOT/.server.pid" ]; then
+    echo "  Stop server:        kill $(cat $PROJECT_ROOT/.server.pid)"
+fi
+echo "  Restart server:     bash scripts/start.sh"
+echo ""
+echo "╔════════════════════════════════════════════════════════════╗"
+echo "║  Server will continue running in the background            ║"
+echo "║  Press Ctrl+C to exit (server will keep running)          ║"
+echo "╚════════════════════════════════════════════════════════════╝"
+echo ""
+
+# Display live logs
+log_info "Displaying server logs (press Ctrl+C to exit)..."
+echo ""
+sleep 1
+
+# Tail logs
+if [ -f "$PROJECT_ROOT/logs/server.log" ]; then
+    tail -f "$PROJECT_ROOT/logs/server.log"
+else
+    log_warning "Log file not found: logs/server.log"
+fi
diff --git a/scripts/retrieve_token.py b/scripts/retrieve_token.py
new file mode 100755
index 0000000..44887a4
--- /dev/null
+++ b/scripts/retrieve_token.py
@@ -0,0 +1,262 @@
+#!/usr/bin/env python3
+"""
+Token Retrieval Script using Playwright
+Automates login to Z.AI and extracts the authentication token
+"""
+import asyncio
+import sys
+import os
+from playwright.async_api import async_playwright, TimeoutError as PlaywrightTimeout
+
+
+async def retrieve_zai_token(email: str, password: str, timeout: int = 60000) -> dict:
+    """
+    Login to Z.AI and retrieve authentication token
+    
+    Args:
+        email: User email for login
+        password: User password for login
+        timeout: Timeout in milliseconds (default: 60s)
+    
+    Returns:
+        dict: Contains 'success' (bool), 'token' (str), 'error' (str)
+    """
+    result = {
+        'success': False,
+        'token': None,
+        'error': None
+    }
+    
+    async with async_playwright() as p:
+        try:
+            # Launch browser in headless mode
+            browser = await p.chromium.launch(
+                headless=True,
+                args=['--no-sandbox', '--disable-setuid-sandbox']
+            )
+            context = await browser.new_context(
+                viewport={'width': 1280, 'height': 720},
+                user_agent='Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36'
+            )
+            page = await context.new_page()
+            
+            print("🌐 Navigating to Z.AI login page...")
+            await page.goto('https://z.ai/', timeout=timeout)
+            await page.wait_for_load_state('networkidle', timeout=30000)
+            
+            # Wait for and click login button
+            print("🔍 Looking for login button...")
+            try:
+                # Try multiple possible selectors for login button
+                selectors = [
+                    'button:has-text("登录")',
+                    'button:has-text("Login")',
+                    'a:has-text("登录")',
+                    'a:has-text("Login")',
+                    '[href*="login"]',
+                    '.login-button',
+                    '#login-btn'
+                ]
+                
+                login_clicked = False
+                for selector in selectors:
+                    try:
+                        await page.click(selector, timeout=5000)
+                        login_clicked = True
+                        print(f"✅ Clicked login button using selector: {selector}")
+                        break
+                    except:
+                        continue
+                
+                if not login_clicked:
+                    raise Exception("Could not find login button")
+                
+                await page.wait_for_load_state('networkidle', timeout=30000)
+            except Exception as e:
+                result['error'] = f"Login button not found: {str(e)}"
+                await browser.close()
+                return result
+            
+            # Enter email
+            print("✉️  Entering email...")
+            email_selectors = [
+                'input[type="email"]',
+                'input[name="email"]',
+                'input[placeholder*="邮箱"]',
+                'input[placeholder*="email"]',
+                '#email'
+            ]
+            
+            email_entered = False
+            for selector in email_selectors:
+                try:
+                    await page.fill(selector, email, timeout=5000)
+                    email_entered = True
+                    print(f"✅ Entered email using selector: {selector}")
+                    break
+                except:
+                    continue
+            
+            if not email_entered:
+                result['error'] = "Email input field not found"
+                await browser.close()
+                return result
+            
+            # Enter password
+            print("🔐 Entering password...")
+            password_selectors = [
+                'input[type="password"]',
+                'input[name="password"]',
+                'input[placeholder*="密码"]',
+                'input[placeholder*="password"]',
+                '#password'
+            ]
+            
+            password_entered = False
+            for selector in password_selectors:
+                try:
+                    await page.fill(selector, password, timeout=5000)
+                    password_entered = True
+                    print(f"✅ Entered password using selector: {selector}")
+                    break
+                except:
+                    continue
+            
+            if not password_entered:
+                result['error'] = "Password input field not found"
+                await browser.close()
+                return result
+            
+            # Submit login form
+            print("📤 Submitting login form...")
+            submit_selectors = [
+                'button[type="submit"]',
+                'button:has-text("登录")',
+                'button:has-text("Login")',
+                '.submit-button',
+                '#submit-btn'
+            ]
+            
+            submitted = False
+            for selector in submit_selectors:
+                try:
+                    await page.click(selector, timeout=5000)
+                    submitted = True
+                    print(f"✅ Clicked submit button using selector: {selector}")
+                    break
+                except:
+                    continue
+            
+            if not submitted:
+                # Try pressing Enter as fallback
+                await page.keyboard.press('Enter')
+            
+            # Wait for navigation after login
+            print("⏳ Waiting for login to complete...")
+            await page.wait_for_load_state('networkidle', timeout=30000)
+            await asyncio.sleep(3)  # Additional wait for tokens to be set
+            
+            # Extract token from localStorage/cookies/network
+            print("🔍 Extracting authentication token...")
+            
+            # Method 1: Check localStorage
+            token = await page.evaluate('''() => {
+                return localStorage.getItem('token') || 
+                       localStorage.getItem('auth_token') ||
+                       localStorage.getItem('access_token') ||
+                       localStorage.getItem('jwt');
+            }''')
+            
+            # Method 2: Check cookies if localStorage didn't work
+            if not token:
+                cookies = await context.cookies()
+                for cookie in cookies:
+                    if cookie['name'] in ['token', 'auth_token', 'access_token', 'jwt', 'session']:
+                        token = cookie['value']
+                        break
+            
+            # Method 3: Check network requests for auth headers
+            if not token:
+                # Make a test API request to trigger auth header
+                try:
+                    response = await page.request.get('https://z.ai/api/config')
+                    headers = response.headers
+                    auth_header = headers.get('authorization', '')
+                    if auth_header.startswith('Bearer '):
+                        token = auth_header.replace('Bearer ', '')
+                except:
+                    pass
+            
+            if token:
+                result['success'] = True
+                result['token'] = token
+                print(f"✅ Token retrieved successfully: {token[:20]}...")
+            else:
+                result['error'] = "Token not found in localStorage, cookies, or network requests"
+                print("❌ Could not extract token from any source")
+            
+            await browser.close()
+            
+        except PlaywrightTimeout as e:
+            result['error'] = f"Timeout during login process: {str(e)}"
+            print(f"⏱️ Timeout: {result['error']}")
+        except Exception as e:
+            result['error'] = f"Unexpected error: {str(e)}"
+            print(f"❌ Error: {result['error']}")
+    
+    return result
+
+
+async def main():
+    """Main entry point"""
+    # Get credentials from environment variables
+    email = os.getenv('ZAI_EMAIL')
+    password = os.getenv('ZAI_PASSWORD')
+    
+    if not email or not password:
+        print("❌ Error: ZAI_EMAIL and ZAI_PASSWORD environment variables must be set")
+        sys.exit(1)
+    
+    print("=" * 60)
+    print("Z.AI Token Retrieval Script")
+    print("=" * 60)
+    print(f"📧 Email: {email}")
+    print(f"🔐 Password: {'*' * len(password)}")
+    print("=" * 60)
+    
+    result = await retrieve_zai_token(email, password)
+    
+    if result['success']:
+        # Write token to .env file
+        env_path = os.path.join(os.path.dirname(os.path.dirname(__file__)), '.env')
+        
+        # Read existing .env or create new one
+        env_content = {}
+        if os.path.exists(env_path):
+            with open(env_path, 'r') as f:
+                for line in f:
+                    line = line.strip()
+                    if line and not line.startswith('#') and '=' in line:
+                        key, value = line.split('=', 1)
+                        env_content[key.strip()] = value.strip()
+        
+        # Update AUTH_TOKEN
+        env_content['AUTH_TOKEN'] = result['token']
+        
+        # Write back to .env
+        with open(env_path, 'w') as f:
+            for key, value in env_content.items():
+                f.write(f"{key}={value}\n")
+        
+        print(f"✅ Token saved to {env_path}")
+        print("=" * 60)
+        sys.exit(0)
+    else:
+        print(f"❌ Failed to retrieve token: {result['error']}")
+        print("=" * 60)
+        sys.exit(1)
+
+
+if __name__ == '__main__':
+    asyncio.run(main())
+
diff --git a/scripts/send_openai_request.sh b/scripts/send_openai_request.sh
new file mode 100755
index 0000000..8d6aad9
--- /dev/null
+++ b/scripts/send_openai_request.sh
@@ -0,0 +1,286 @@
+#!/bin/bash
+#
+# send_openai_request.sh - Test Z.AI2API with REAL OpenAI-compatible requests
+#
+# This script makes ACTUAL API calls to your running z.ai2api_python server,
+# which proxies to the real Z.AI chat interface using your JWT token.
+# 
+# NO MOCKS - All responses are real AI-generated content from Z.AI!
+#
+# The server:
+#  ✅ Mimics a web browser accessing Z.AI's chat interface
+#  ✅ Uses the JWT token from your browser session
+#  ✅ Makes requests exactly like the web UI does
+#  ✅ Converts Z.AI's web responses to OpenAI API format
+#
+
+set -e  # Exit on error
+
+# Colors for output
+RED='\033[0;31m'
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+BLUE='\033[0;34m'
+CYAN='\033[0;36m'
+NC='\033[0m' # No Color
+
+# Script directory
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+PROJECT_ROOT="$(dirname "$SCRIPT_DIR")"
+
+# Logging functions
+log_info() {
+    echo -e "${BLUE}ℹ️  $1${NC}"
+}
+
+log_success() {
+    echo -e "${GREEN}✅ $1${NC}"
+}
+
+log_warning() {
+    echo -e "${YELLOW}⚠️  $1${NC}"
+}
+
+log_error() {
+    echo -e "${RED}❌ $1${NC}"
+}
+
+# Banner
+echo ""
+echo "============================================================"
+echo "       Z.AI2API - OpenAI API Request Test"
+echo "============================================================"
+echo ""
+
+# Navigate to project root
+cd "$PROJECT_ROOT"
+
+# Load environment variables if .env exists
+if [ -f ".env" ]; then
+    set -a
+    source .env
+    set +a
+fi
+
+# Configuration
+PORT=${LISTEN_PORT:-8080}
+API_URL="http://localhost:$PORT/v1/chat/completions"
+AUTH_TOKEN=${AUTH_TOKEN:-"sk-test"}
+
+# Step 1: Check if server is running
+log_info "Checking if server is running on port $PORT..."
+if ! curl -s -f "http://localhost:$PORT/" > /dev/null 2>&1; then
+    log_error "Server is not responding on port $PORT"
+    log_info "Please start the server first: bash scripts/start.sh"
+    exit 1
+fi
+log_success "Server is responding"
+
+# Step 2: Prepare test request
+log_info "Preparing test request..."
+
+# Test message - Real question about Graph-RAG
+TEST_MESSAGE="What is Graph-RAG?"
+
+# Create request payload
+REQUEST_PAYLOAD=$(cat <<EOF
+{
+  "model": "GLM-4-6-API-V1",
+  "messages": [
+    {
+      "role": "user",
+      "content": "$TEST_MESSAGE"
+    }
+  ],
+  "stream": false,
+  "temperature": 0.7,
+  "max_tokens": 500
+}
+EOF
+)
+
+log_success "Request prepared"
+
+# Step 3: Send request
+echo ""
+echo "============================================================"
+echo "                 Sending API Request..."
+echo "============================================================"
+echo ""
+echo -e "${CYAN}Request Details:${NC}"
+echo "  URL:     $API_URL"
+echo "  Model:   GLM-4-6-API-V1"
+echo "  Message: $TEST_MESSAGE"
+echo ""
+log_info "Sending request..."
+
+# Send request and capture response
+RESPONSE=$(curl -s -w "\n%{http_code}" \
+  -X POST "$API_URL" \
+  -H "Content-Type: application/json" \
+  -H "Authorization: Bearer $AUTH_TOKEN" \
+  -d "$REQUEST_PAYLOAD")
+
+# Extract HTTP status code (last line)
+HTTP_CODE=$(echo "$RESPONSE" | tail -n 1)
+RESPONSE_BODY=$(echo "$RESPONSE" | head -n -1)
+
+# Step 4: Display response
+echo ""
+echo "============================================================"
+echo "                   API Response"
+echo "============================================================"
+echo ""
+
+if [ "$HTTP_CODE" -eq 200 ]; then
+    log_success "Request successful (HTTP $HTTP_CODE)"
+    echo ""
+    
+    # Parse and display response using Python for better formatting
+    if command -v python3 &> /dev/null; then
+        echo -e "${CYAN}Response:${NC}"
+        echo "$RESPONSE_BODY" | python3 -c "
+import sys
+import json
+
+try:
+    data = json.load(sys.stdin)
+    
+    # Display basic info
+    print(f\"  Model: {data.get('model', 'N/A')}\")
+    print(f\"  ID: {data.get('id', 'N/A')}\")
+    print(f\"  Created: {data.get('created', 'N/A')}\")
+    print()
+    
+    # Display message content
+    if 'choices' in data and len(data['choices']) > 0:
+        choice = data['choices'][0]
+        message = choice.get('message', {})
+        content = message.get('content', 'No content')
+        finish_reason = choice.get('finish_reason', 'N/A')
+        
+        print('  Content:')
+        print('  ' + '-' * 58)
+        for line in content.split('\n'):
+            print(f'  {line}')
+        print('  ' + '-' * 58)
+        print()
+        print(f\"  Finish Reason: {finish_reason}\")
+    
+    # Display usage info
+    if 'usage' in data:
+        usage = data['usage']
+        print()
+        print('  Token Usage:')
+        print(f\"    Prompt: {usage.get('prompt_tokens', 'N/A')} tokens\")
+        print(f\"    Completion: {usage.get('completion_tokens', 'N/A')} tokens\")
+        print(f\"    Total: {usage.get('total_tokens', 'N/A')} tokens\")
+    
+    print()
+    
+except json.JSONDecodeError:
+    print('  [Raw Response]')
+    print(sys.stdin.read())
+except Exception as e:
+    print(f'  Error parsing response: {e}')
+    print()
+    print('  [Raw Response]')
+    print(sys.stdin.read())
+"
+    else
+        # Fallback: display raw JSON
+        echo "$RESPONSE_BODY" | head -c 1000
+        echo ""
+    fi
+    
+    echo ""
+    echo "============================================================"
+    echo ""
+    log_success "API test completed successfully! ✨"
+    
+else
+    log_error "Request failed (HTTP $HTTP_CODE)"
+    echo ""
+    echo -e "${CYAN}Error Response:${NC}"
+    echo "$RESPONSE_BODY" | head -c 500
+    echo ""
+    echo ""
+    echo "============================================================"
+    echo ""
+    log_error "API test failed"
+    
+    if [ "$HTTP_CODE" -eq 401 ] || [ "$HTTP_CODE" -eq 403 ]; then
+        log_warning "Authentication issue - check AUTH_TOKEN in .env"
+    elif [ "$HTTP_CODE" -eq 500 ]; then
+        log_warning "Server error - check logs/server.log"
+    fi
+    
+    exit 1
+fi
+
+# Step 5: Additional test with streaming
+echo ""
+log_info "Testing streaming mode..."
+echo ""
+
+STREAM_REQUEST=$(cat <<EOF
+{
+  "model": "GLM-4-6-API-V1",
+  "messages": [
+    {
+      "role": "user",
+      "content": "Briefly explain the key benefits of Graph-RAG in 2-3 sentences."
+    }
+  ],
+  "stream": true,
+  "max_tokens": 200
+}
+EOF
+)
+
+echo -e "${CYAN}Streaming Response:${NC}"
+echo "------------------------------------------------------------"
+curl -s -N \
+  -X POST "$API_URL" \
+  -H "Content-Type: application/json" \
+  -H "Authorization: Bearer $AUTH_TOKEN" \
+  -d "$STREAM_REQUEST" | while IFS= read -r line; do
+    if [[ $line == data:* ]]; then
+        # Extract JSON after "data: "
+        json_data="${line#data: }"
+        if [ "$json_data" != "[DONE]" ]; then
+            # Parse and display delta content
+            echo "$json_data" | python3 -c "
+import sys
+import json
+try:
+    data = json.load(sys.stdin)
+    if 'choices' in data and len(data['choices']) > 0:
+        delta = data['choices'][0].get('delta', {})
+        content = delta.get('content', '')
+        if content:
+            print(content, end='', flush=True)
+except:
+    pass
+" 2>/dev/null
+        fi
+    fi
+done
+echo ""
+echo "------------------------------------------------------------"
+echo ""
+log_success "Streaming test completed! ✨"
+
+# Final summary
+echo ""
+echo "============================================================"
+echo "              API Testing Complete! ✅"
+echo "============================================================"
+echo ""
+echo "🎉 All tests passed successfully!"
+echo ""
+echo "📖 API Documentation: http://localhost:$PORT/docs"
+echo "🔧 Admin Panel: http://localhost:$PORT/admin"
+echo ""
+echo "============================================================"
+echo ""
diff --git a/scripts/setup.sh b/scripts/setup.sh
new file mode 100755
index 0000000..c1e6ad0
--- /dev/null
+++ b/scripts/setup.sh
@@ -0,0 +1,289 @@
+#!/bin/bash
+#
+# setup.sh - Z.AI2API Setup Script
+# 
+# This script performs initial setup:
+# 1. Installs Python dependencies
+# 2. Installs Playwright and browsers
+# 3. Creates .env file if needed
+# 4. Retrieves authentication token via Playwright login
+# 5. Initializes the database
+#
+
+set -e  # Exit on error
+
+# Colors for output
+RED='\033[0;31m'
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+BLUE='\033[0;34m'
+NC='\033[0m' # No Color
+
+# Script directory
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+PROJECT_ROOT="$(dirname "$SCRIPT_DIR")"
+
+# Logging functions
+log_info() {
+    echo -e "${BLUE}ℹ️  $1${NC}"
+}
+
+log_success() {
+    echo -e "${GREEN}✅ $1${NC}"
+}
+
+log_warning() {
+    echo -e "${YELLOW}⚠️  $1${NC}"
+}
+
+log_error() {
+    echo -e "${RED}❌ $1${NC}"
+}
+
+# Banner
+echo ""
+echo "============================================================"
+echo "         Z.AI2API Python - Setup Script"
+echo "============================================================"
+echo ""
+
+# Check for required environment variables
+log_info "Checking environment variables..."
+if [ -z "$ZAI_EMAIL" ] || [ -z "$ZAI_PASSWORD" ]; then
+    log_error "ZAI_EMAIL and ZAI_PASSWORD environment variables must be set"
+    echo ""
+    echo "Usage:"
+    echo "  export ZAI_EMAIL=your-email@example.com"
+    echo "  export ZAI_PASSWORD=your-password"
+    echo "  bash scripts/setup.sh"
+    echo ""
+    exit 1
+fi
+log_success "Environment variables found"
+
+# Navigate to project root
+cd "$PROJECT_ROOT"
+log_info "Working directory: $PROJECT_ROOT"
+
+# Step 1: Find suitable Python version (3.9-3.12)
+log_info "Finding suitable Python version (3.9-3.12)..."
+
+SUITABLE_PYTHON=""
+SUITABLE_VERSION=""
+
+# Try different Python versions in order of preference (avoid 3.13 due to pydantic issues)
+for py_cmd in python3.11 python3.12 python3.10 python3.9 python3; do
+    if command -v $py_cmd &> /dev/null; then
+        version=$($py_cmd --version 2>&1 | cut -d' ' -f2)
+        major=$(echo $version | cut -d'.' -f1)
+        minor=$(echo $version | cut -d'.' -f2)
+        
+        # Check if version is 3.9-3.12 (exclude 3.13+)
+        if [ "$major" -eq 3 ] && [ "$minor" -ge 9 ] && [ "$minor" -le 12 ]; then
+            SUITABLE_PYTHON=$py_cmd
+            SUITABLE_VERSION=$version
+            break
+        fi
+    fi
+done
+
+if [ -z "$SUITABLE_PYTHON" ]; then
+    log_error "No suitable Python version found!"
+    echo ""
+    echo "Python 3.9-3.12 is required (Python 3.13+ has compatibility issues)"
+    echo ""
+    echo "Please install one of these Python versions:"
+    echo "  - Python 3.11 (recommended)"
+    echo "  - Python 3.12"
+    echo "  - Python 3.10"
+    echo "  - Python 3.9"
+    exit 1
+fi
+
+log_success "Found suitable Python: $SUITABLE_PYTHON ($SUITABLE_VERSION)"
+
+# Step 1b: Create virtual environment if it doesn't exist
+VENV_DIR="$PROJECT_ROOT/.venv"
+if [ ! -d "$VENV_DIR" ]; then
+    log_info "Creating virtual environment with $SUITABLE_PYTHON..."
+    
+    if command -v uv &> /dev/null; then
+        # Use uv to create venv (faster)
+        uv venv --python $SUITABLE_PYTHON "$VENV_DIR"
+    else
+        # Use standard Python venv
+        $SUITABLE_PYTHON -m venv "$VENV_DIR"
+    fi
+    
+    if [ $? -eq 0 ]; then
+        log_success "Virtual environment created at $VENV_DIR"
+    else
+        log_error "Failed to create virtual environment"
+        exit 1
+    fi
+else
+    log_success "Virtual environment already exists"
+fi
+
+# Activate virtual environment
+log_info "Activating virtual environment..."
+source "$VENV_DIR/bin/activate"
+
+if [ $? -eq 0 ]; then
+    log_success "Virtual environment activated"
+    log_info "Python: $(which python)"
+else
+    log_error "Failed to activate virtual environment"
+    exit 1
+fi
+
+# Step 2: Install Python dependencies
+log_info "Installing Python dependencies..."
+if command -v uv &> /dev/null; then
+    log_info "Using uv for dependency management..."
+    uv sync
+    PIP_CMD="uv pip install"
+else
+    log_info "Using pip for dependency management..."
+    python3 -m pip install --upgrade pip
+    python3 -m pip install -r requirements.txt
+    PIP_CMD="python3 -m pip install"
+fi
+log_success "Python dependencies installed"
+
+# Step 3: Install Playwright
+log_info "Installing Playwright..."
+
+# Determine which Python command to use
+if command -v uv &> /dev/null; then
+    PYTHON_CMD="uv run python"
+    PLAYWRIGHT_CMD="uv run playwright"
+else
+    PYTHON_CMD="python3"
+    PLAYWRIGHT_CMD="python3 -m playwright"
+fi
+
+if ! $PYTHON_CMD -c "import playwright" 2>/dev/null; then
+    log_info "Playwright not found, installing..."
+    $PIP_CMD playwright
+fi
+
+# Install Playwright browsers
+log_info "Installing Playwright browsers (this may take a few minutes)..."
+$PLAYWRIGHT_CMD install chromium
+log_success "Playwright and browsers installed"
+
+# Step 4: Create .env file if it doesn't exist
+log_info "Checking .env file..."
+if [ ! -f ".env" ]; then
+    if [ -f ".env.example" ]; then
+        log_info "Creating .env from .env.example..."
+        cp .env.example .env
+        log_success ".env file created"
+    else
+        log_warning ".env.example not found, creating minimal .env..."
+        cat > .env << EOF
+# Z.AI2API Configuration
+AUTH_TOKEN=sk-placeholder
+SKIP_AUTH_TOKEN=false
+TOKEN_FAILURE_THRESHOLD=3
+TOKEN_RECOVERY_TIMEOUT=1800
+ANONYMOUS_MODE=true
+LISTEN_PORT=8080
+SERVICE_NAME=z-ai2api-server
+DEBUG_LOGGING=false
+TOOL_SUPPORT=true
+SCAN_LIMIT=200000
+EOF
+        log_success "Minimal .env file created"
+    fi
+else
+    log_success ".env file already exists"
+fi
+
+# Step 5: Retrieve authentication token
+log_info "Retrieving authentication token from Z.AI..."
+echo ""
+echo "============================================================"
+echo "              Token Retrieval (Playwright)"
+echo "============================================================"
+echo ""
+
+if command -v uv &> /dev/null; then
+    uv run python scripts/retrieve_token.py
+    TOKEN_EXIT_CODE=$?
+else
+    python3 scripts/retrieve_token.py
+    TOKEN_EXIT_CODE=$?
+fi
+
+echo ""
+if [ $TOKEN_EXIT_CODE -eq 0 ]; then
+    log_success "Token retrieved and saved to .env"
+else
+    log_error "Failed to retrieve token"
+    log_warning "You can manually set AUTH_TOKEN in .env file"
+    exit 1
+fi
+
+# Step 6: Create logs directory
+log_info "Creating logs directory..."
+mkdir -p logs
+log_success "Logs directory ready"
+
+# Step 7: Test database initialization
+log_info "Testing database initialization..."
+if command -v uv &> /dev/null; then
+    uv run python -c "
+import asyncio
+from app.services.token_dao import init_token_database
+
+async def test():
+    await init_token_database()
+    print('Database initialized successfully')
+
+asyncio.run(test())
+" 2>&1 | tail -1
+else
+    python3 -c "
+import asyncio
+from app.services.token_dao import init_token_database
+
+async def test():
+    await init_token_database()
+    print('Database initialized successfully')
+
+asyncio.run(test())
+" 2>&1 | tail -1
+fi
+
+if [ $? -eq 0 ]; then
+    log_success "Database initialization test passed"
+else
+    log_warning "Database will be initialized on first server start"
+fi
+
+# Final summary
+echo ""
+echo "============================================================"
+echo "                   Setup Complete! ✨"
+echo "============================================================"
+echo ""
+log_success "Virtual environment is active!"
+log_info "To activate in new terminals, run:"
+echo ""
+echo "    source .venv/bin/activate"
+echo ""
+echo "📋 Next steps:"
+echo "  1. Review .env file and adjust settings if needed"
+echo "  2. Run: bash scripts/start.sh"
+echo "  3. Or run everything at once: bash scripts/all.sh"
+echo ""
+echo "🌐 Server will be available at: http://localhost:8080"
+echo "📖 API Documentation: http://localhost:8080/docs"
+echo "🔧 Admin Panel: http://localhost:8080/admin"
+echo ""
+echo "💡 Note: The virtual environment will remain active in this shell"
+echo ""
+echo "============================================================"
+echo ""
diff --git a/scripts/start.sh b/scripts/start.sh
new file mode 100755
index 0000000..37e0672
--- /dev/null
+++ b/scripts/start.sh
@@ -0,0 +1,218 @@
+#!/bin/bash
+#
+# start.sh - Z.AI2API Server Startup Script
+#
+# This script starts the Z.AI2API server using the AUTH_TOKEN from .env
+# The token was retrieved via Playwright automation in setup.sh
+#
+# The server will:
+#  ✅ Load AUTH_TOKEN from .env file
+#  ✅ Use it to authenticate with Z.AI's chat interface
+#  ✅ Proxy all OpenAI API requests to real Z.AI service
+#  ✅ Return actual AI-generated responses
+#
+
+set -e  # Exit on error
+
+# Colors for output
+RED='\033[0;31m'
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+BLUE='\033[0;34m'
+NC='\033[0m' # No Color
+
+# Script directory
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+PROJECT_ROOT="$(dirname "$SCRIPT_DIR")"
+PID_FILE="$PROJECT_ROOT/.server.pid"
+
+# Logging functions
+log_info() {
+    echo -e "${BLUE}ℹ️  $1${NC}"
+}
+
+log_success() {
+    echo -e "${GREEN}✅ $1${NC}"
+}
+
+log_warning() {
+    echo -e "${YELLOW}⚠️  $1${NC}"
+}
+
+log_error() {
+    echo -e "${RED}❌ $1${NC}"
+}
+
+# Banner
+echo ""
+echo "============================================================"
+echo "         Z.AI2API Python - Server Startup"
+echo "============================================================"
+echo ""
+
+# Navigate to project root
+cd "$PROJECT_ROOT"
+
+# Step 0: Activate virtual environment if it exists
+VENV_DIR="$PROJECT_ROOT/.venv"
+if [ -d "$VENV_DIR" ]; then
+    log_info "Activating virtual environment..."
+    source "$VENV_DIR/bin/activate"
+    if [ $? -eq 0 ]; then
+        log_success "Virtual environment activated"
+    else
+        log_warning "Failed to activate virtual environment, using system Python"
+    fi
+else
+    log_warning "Virtual environment not found, using system Python"
+    log_info "Run 'bash scripts/setup.sh' to create one"
+fi
+
+# Step 1: Check if .env file exists
+log_info "Checking configuration..."
+if [ ! -f ".env" ]; then
+    log_error ".env file not found"
+    log_info "Please run: bash scripts/setup.sh"
+    exit 1
+fi
+log_success "Configuration file found"
+
+# Load environment variables
+set -a
+source .env
+set +a
+
+# Verify AUTH_TOKEN is loaded
+if [ -n "$AUTH_TOKEN" ]; then
+    TOKEN_PREVIEW="${AUTH_TOKEN:0:10}..."
+    log_success "AUTH_TOKEN loaded: $TOKEN_PREVIEW"
+else
+    log_warning "AUTH_TOKEN not found in .env"
+fi
+
+# Get port from .env or use default
+PORT=${LISTEN_PORT:-8080}
+
+# Step 2: Check if port is available
+log_info "Checking if port $PORT is available..."
+if lsof -Pi :$PORT -sTCP:LISTEN -t >/dev/null 2>&1; then
+    PID=$(lsof -Pi :$PORT -sTCP:LISTEN -t)
+    log_warning "Port $PORT is already in use by PID $PID"
+    
+    # Check if it's our server
+    if [ -f "$PID_FILE" ] && [ "$(cat $PID_FILE)" == "$PID" ]; then
+        log_info "Server is already running (PID: $PID)"
+        log_success "Server is healthy at http://localhost:$PORT"
+        exit 0
+    else
+        log_error "Another process is using port $PORT"
+        log_info "Please stop the process or change LISTEN_PORT in .env"
+        exit 1
+    fi
+fi
+log_success "Port $PORT is available"
+
+# Step 3: Check Python dependencies
+log_info "Checking dependencies..."
+if ! python3 -c "import fastapi, granian, httpx" 2>/dev/null; then
+    log_error "Required dependencies not found"
+    log_info "Please run: bash scripts/setup.sh"
+    exit 1
+fi
+log_success "Dependencies verified"
+
+# Step 4: Start the server
+log_info "Starting server..."
+echo ""
+echo "============================================================"
+echo "                  Server Starting..."
+echo "============================================================"
+echo ""
+
+# Determine which Python command to use
+if command -v uv &> /dev/null; then
+    PYTHON_CMD="uv run python"
+    log_info "Using uv runtime"
+else
+    PYTHON_CMD="python3"
+    log_info "Using system Python"
+fi
+
+# Start server in background
+nohup $PYTHON_CMD main.py > logs/server.log 2>&1 &
+SERVER_PID=$!
+
+# Save PID to file
+echo $SERVER_PID > "$PID_FILE"
+log_success "Server started with PID: $SERVER_PID"
+
+# Step 5: Wait for server to be ready
+log_info "Waiting for server to be ready..."
+MAX_ATTEMPTS=30
+ATTEMPT=0
+
+while [ $ATTEMPT -lt $MAX_ATTEMPTS ]; do
+    if curl -s -f "http://localhost:$PORT/" > /dev/null 2>&1; then
+        log_success "Server is ready!"
+        break
+    fi
+    
+    # Check if process is still running
+    if ! kill -0 $SERVER_PID 2>/dev/null; then
+        log_error "Server process died unexpectedly"
+        log_info "Check logs/server.log for details"
+        rm -f "$PID_FILE"
+        exit 1
+    fi
+    
+    ATTEMPT=$((ATTEMPT + 1))
+    echo -n "."
+    sleep 1
+done
+
+echo ""
+
+if [ $ATTEMPT -eq $MAX_ATTEMPTS ]; then
+    log_error "Server failed to start within ${MAX_ATTEMPTS} seconds"
+    log_info "Check logs/server.log for details"
+    kill $SERVER_PID 2>/dev/null || true
+    rm -f "$PID_FILE"
+    exit 1
+fi
+
+# Step 6: Display server information
+echo ""
+echo "============================================================"
+echo "             Server Started Successfully! 🚀"
+echo "============================================================"
+echo ""
+echo "📊 Server Information:"
+echo "  PID:        $SERVER_PID"
+echo "  Port:       $PORT"
+echo "  Log file:   logs/server.log"
+echo "  PID file:   $PID_FILE"
+echo ""
+echo "🌐 Endpoints:"
+echo "  API Root:   http://localhost:$PORT/"
+echo "  API Docs:   http://localhost:$PORT/docs"
+echo "  Admin:      http://localhost:$PORT/admin"
+echo ""
+echo "📝 Commands:"
+echo "  View logs:  tail -f logs/server.log"
+echo "  Stop:       kill $SERVER_PID  (or: kill \$(cat $PID_FILE))"
+echo "  Test API:   bash scripts/send_openai_request.sh"
+echo ""
+echo "============================================================"
+echo ""
+
+# Keep script running to show initial logs
+log_info "Showing initial logs (press Ctrl+C to exit, server will keep running):"
+echo ""
+tail -f logs/server.log &
+TAIL_PID=$!
+
+# Trap Ctrl+C to stop tail but keep server running
+trap "kill $TAIL_PID 2>/dev/null; echo ''; log_success 'Server is running in background'; exit 0" INT
+
+# Wait for tail to finish (won't happen unless killed)
+wait $TAIL_PID