diff --git a/README-CN.md b/README-CN.md
new file mode 100644
index 0000000..12eec25
--- /dev/null
+++ b/README-CN.md
@@ -0,0 +1,576 @@
+# 🎬 Huobao Drama - AI 短剧生成平台
+
+<div align="center">
+
+**基于 Go + Vue3 的全栈 AI 短剧自动化生产平台**
+
+[![Go Version](https://img.shields.io/badge/Go-1.23+-00ADD8?style=flat&logo=go)](https://golang.org)
+[![Vue Version](https://img.shields.io/badge/Vue-3.x-4FC08D?style=flat&logo=vue.js)](https://vuejs.org)
+[![License](https://img.shields.io/badge/License-CC%20BY--NC--SA%204.0-lightgrey.svg)](https://creativecommons.org/licenses/by-nc-sa/4.0/)
+
+[功能特性](#功能特性) • [快速开始](#快速开始) • [部署指南](#部署指南)
+
+[简体中文](README-CN.md) | [English](README.md) | [日本語](README-JA.md)
+
+</div>
+
+---
+
+## 📖 项目简介
+
+Huobao Drama 是一个基于 AI 的短剧自动化生产平台，实现从剧本生成、角色设计、分镜制作到视频合成的全流程自动化。
+
+### 🎯 核心价值
+
+- **🤖 AI 驱动**：使用大语言模型解析剧本，提取角色、场景和分镜信息
+- **🎨 智能创作**：AI 绘图生成角色形象和场景背景
+- **📹 视频生成**：基于文生视频和图生视频模型自动生成分镜视频
+- **🔄 工作流**：完整的短剧制作工作流，从创意到成片一站式完成
+
+### 🛠️ 技术架构
+
+采用**DDD 领域驱动设计**，清晰分层：
+
+```
+├── API层 (Gin HTTP)
+├── 应用服务层 (Business Logic)
+├── 领域层 (Domain Models)
+└── 基础设施层 (Database, External Services)
+```
+
+### 🎥 作品展示 / Demo Videos
+
+体验 AI 短剧生成效果：
+
+<div align="center">
+
+**示例作品 1**
+
+<video src="https://ffile.chatfire.site/cf/public/20260114094337396.mp4" controls width="640"></video>
+
+**示例作品 2**
+
+<video src="https://ffile.chatfire.site/cf/public/fcede75e8aeafe22031dbf78f86285b8.mp4" controls width="640"></video>
+
+[点击观看视频 1](https://ffile.chatfire.site/cf/public/20260114094337396.mp4) | [点击观看视频 2](https://ffile.chatfire.site/cf/public/fcede75e8aeafe22031dbf78f86285b8.mp4)
+
+</div>
+
+---
+
+## ✨ 功能特性
+
+### 🎭 角色管理
+
+- ✅ AI 生成角色形象
+- ✅ 批量角色生成
+- ✅ 角色图片上传和管理
+
+### 🎬 分镜制作
+
+- ✅ 自动生成分镜脚本
+- ✅ 场景描述和镜头设计
+- ✅ 分镜图片生成（文生图）
+- ✅ 帧类型选择（首帧/关键帧/尾帧/分镜板）
+
+### 🎥 视频生成
+
+- ✅ 图生视频自动生成
+- ✅ 视频合成和剪辑
+- ✅ 转场效果
+
+### 📦 资源管理
+
+- ✅ 素材库统一管理
+- ✅ 本地存储支持
+- ✅ 资源导入导出
+- ✅ 任务进度追踪
+
+---
+
+## 🚀 快速开始
+
+### 📋 环境要求
+
+| 软件        | 版本要求 | 说明                 |
+| ----------- | -------- | -------------------- |
+| **Go**      | 1.23+    | 后端运行环境         |
+| **Node.js** | 18+      | 前端构建环境         |
+| **npm**     | 9+       | 包管理工具           |
+| **FFmpeg**  | 4.0+     | 视频处理（**必需**） |
+| **SQLite**  | 3.x      | 数据库（已内置）     |
+
+#### 安装 FFmpeg
+
+**macOS:**
+
+```bash
+brew install ffmpeg
+```
+
+**Ubuntu/Debian:**
+
+```bash
+sudo apt update
+sudo apt install ffmpeg
+```
+
+**Windows:**
+从 [FFmpeg 官网](https://ffmpeg.org/download.html) 下载并配置环境变量
+
+验证安装：
+
+```bash
+ffmpeg -version
+```
+
+### ⚙️ 配置文件
+
+复制并编辑配置文件：
+
+```bash
+cp configs/config.example.yaml configs/config.yaml
+vim configs/config.yaml
+```
+
+配置文件格式（`configs/config.yaml`）：
+
+```yaml
+app:
+  name: "Huobao Drama API"
+  version: "1.0.0"
+  debug: true # 开发环境设为true，生产环境设为false
+
+server:
+  port: 5678
+  host: "0.0.0.0"
+  cors_origins:
+    - "http://localhost:3012"
+  read_timeout: 600
+  write_timeout: 600
+
+database:
+  type: "sqlite"
+  path: "./data/drama_generator.db"
+  max_idle: 10
+  max_open: 100
+
+storage:
+  type: "local"
+  local_path: "./data/storage"
+  base_url: "http://localhost:5678/static"
+
+ai:
+  default_text_provider: "openai"
+  default_image_provider: "openai"
+  default_video_provider: "doubao"
+```
+
+**重要配置项：**
+
+- `app.debug`: 调试模式开关（开发环境建议设为 true）
+- `server.port`: 服务运行端口
+- `server.cors_origins`: 允许跨域访问的前端地址
+- `database.path`: SQLite 数据库文件路径
+- `storage.local_path`: 本地文件存储路径
+- `storage.base_url`: 静态资源访问 URL
+- `ai.default_*_provider`: AI 服务提供商配置（在 Web 界面中配置具体的 API Key）
+
+### 📥 安装依赖
+
+```bash
+# 克隆项目
+git clone https://github.com/chatfire-AI/huobao-drama.git
+cd huobao-drama
+
+# 安装Go依赖
+go mod download
+
+# 安装前端依赖
+cd web
+npm install
+cd ..
+```
+
+### 🎯 启动项目
+
+#### 方式一：开发模式（推荐）
+
+**前后端分离，支持热重载**
+
+```bash
+# 终端1：启动后端服务
+go run main.go
+
+# 终端2：启动前端开发服务器
+cd web
+npm run dev
+```
+
+- 前端地址: `http://localhost:3012`
+- 后端 API: `http://localhost:5678/api/v1`
+- 前端自动代理 API 请求到后端
+
+#### 方式二：单服务模式
+
+**后端同时提供 API 和前端静态文件**
+
+```bash
+# 1. 构建前端
+cd web
+npm run build
+cd ..
+
+# 2. 启动服务
+go run main.go
+```
+
+访问: `http://localhost:5678`
+
+### 🗄️ 数据库初始化
+
+数据库表会在首次启动时自动创建（使用 GORM AutoMigrate），无需手动迁移。
+
+---
+
+## 📦 部署指南
+
+### 🐳 Docker 部署（推荐）
+
+#### 方式一：Docker Compose（推荐）
+
+```bash
+# 启动服务
+docker-compose up -d
+
+# 查看日志
+docker-compose logs -f
+
+# 停止服务
+docker-compose down
+```
+
+#### 方式二：Docker 命令
+
+> **注意**：Linux 用户需添加 `--add-host=host.docker.internal:host-gateway` 以访问宿主机服务
+
+```bash
+# 从 Docker Hub 运行
+docker run -d \
+  --name huobao-drama \
+  -p 5678:5678 \
+  -v $(pwd)/data:/app/data \
+  --restart unless-stopped \
+  huobao/huobao-drama:latest
+
+# 查看日志
+docker logs -f huobao-drama
+```
+
+**本地构建**（可选）：
+
+```bash
+docker build -t huobao-drama:latest .
+docker run -d --name huobao-drama -p 5678:5678 -v $(pwd)/data:/app/data huobao-drama:latest
+```
+
+**Docker 部署优势：**
+
+- ✅ 开箱即用，内置默认配置
+- ✅ 环境一致性，避免依赖问题
+- ✅ 一键启动，无需安装 Go、Node.js、FFmpeg
+- ✅ 易于迁移和扩展
+- ✅ 自动健康检查和重启
+- ✅ 自动处理文件权限，无需手动配置
+
+#### 🔗 访问宿主机服务（Ollama/本地模型）
+
+容器已配置支持访问宿主机服务，直接使用 `http://host.docker.internal:端口号` 即可。
+
+**配置步骤：**
+
+1. **宿主机启动服务（监听所有接口）**
+
+   ```bash
+   export OLLAMA_HOST=0.0.0.0:11434 && ollama serve
+   ```
+
+2. **前端 AI 服务配置**
+   - Base URL: `http://host.docker.internal:11434/v1`
+   - Provider: `openai`
+   - Model: `qwen2.5:latest`
+
+---
+
+### 🏭 传统部署方式
+
+#### 1. 编译构建
+
+```bash
+# 1. 构建前端
+cd web
+npm run build
+cd ..
+
+# 2. 编译后端
+go build -o huobao-drama .
+```
+
+生成文件：
+
+- `huobao-drama` - 后端可执行文件
+- `web/dist/` - 前端静态文件（已嵌入后端）
+
+#### 2. 准备部署文件
+
+需要上传到服务器的文件：
+
+```
+huobao-drama            # 后端可执行文件
+configs/config.yaml     # 配置文件
+data/                   # 数据目录（可选，首次运行自动创建）
+```
+
+#### 3. 服务器配置
+
+```bash
+# 上传文件到服务器
+scp huobao-drama user@server:/opt/huobao-drama/
+scp configs/config.yaml user@server:/opt/huobao-drama/configs/
+
+# SSH登录服务器
+ssh user@server
+
+# 修改配置文件
+cd /opt/huobao-drama
+vim configs/config.yaml
+# 设置mode为production
+# 配置域名和存储路径
+
+# 创建数据目录并设置权限（重要！）
+# 注意：将 YOUR_USER 替换为实际运行服务的用户名（如 www-data、ubuntu、deploy 等）
+sudo mkdir -p /opt/huobao-drama/data/storage
+sudo chown -R YOUR_USER:YOUR_USER /opt/huobao-drama/data
+sudo chmod -R 755 /opt/huobao-drama/data
+
+# 赋予执行权限
+chmod +x huobao-drama
+
+# 启动服务
+./huobao-drama
+```
+
+#### 4. 使用 systemd 管理服务
+
+创建服务文件 `/etc/systemd/system/huobao-drama.service`:
+
+```ini
+[Unit]
+Description=Huobao Drama Service
+After=network.target
+
+[Service]
+Type=simple
+User=YOUR_USER
+WorkingDirectory=/opt/huobao-drama
+ExecStart=/opt/huobao-drama/huobao-drama
+Restart=on-failure
+RestartSec=10
+
+# 环境变量（可选）
+# Environment="GIN_MODE=release"
+
+[Install]
+WantedBy=multi-user.target
+```
+
+启动服务：
+
+```bash
+sudo systemctl daemon-reload
+sudo systemctl enable huobao-drama
+sudo systemctl start huobao-drama
+sudo systemctl status huobao-drama
+```
+
+**⚠️ 常见问题：SQLite 写权限错误**
+
+如果遇到 `attempt to write a readonly database` 错误：
+
+```bash
+# 1. 确认当前运行服务的用户
+sudo systemctl status huobao-drama | grep "Main PID"
+ps aux | grep huobao-drama
+
+# 2. 修复权限（将 YOUR_USER 替换为实际用户名）
+sudo chown -R YOUR_USER:YOUR_USER /opt/huobao-drama/data
+sudo chmod -R 755 /opt/huobao-drama/data
+
+# 3. 验证权限
+ls -la /opt/huobao-drama/data
+# 应该显示所有者为运行服务的用户
+
+# 4. 重启服务
+sudo systemctl restart huobao-drama
+```
+
+**原因说明**：
+
+- SQLite 需要对数据库文件 **和** 所在目录都有写权限
+- 需要在目录中创建临时文件（如 `-wal`、`-journal`）
+- **关键**：确保 systemd 配置中的 `User` 与数据目录所有者一致
+
+**常用用户名**：
+
+- Ubuntu/Debian: `www-data`、`ubuntu`
+- CentOS/RHEL: `nginx`、`apache`
+- 自定义部署: `deploy`、`app`、当前登录用户
+
+#### 5. Nginx 反向代理
+
+```nginx
+server {
+    listen 80;
+    server_name your-domain.com;
+
+    location / {
+        proxy_pass http://localhost:5678;
+        proxy_set_header Host $host;
+        proxy_set_header X-Real-IP $remote_addr;
+        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+    }
+
+    # 静态文件直接访问
+    location /static/ {
+        alias /opt/huobao-drama/data/storage/;
+    }
+}
+```
+
+---
+
+## 🎨 技术栈
+
+### 后端技术
+
+- **语言**: Go 1.23+
+- **Web 框架**: Gin 1.9+
+- **ORM**: GORM
+- **数据库**: SQLite
+- **日志**: Zap
+- **视频处理**: FFmpeg
+- **AI 服务**: OpenAI、Gemini、火山等
+
+### 前端技术
+
+- **框架**: Vue 3.4+
+- **语言**: TypeScript 5+
+- **构建工具**: Vite 5
+- **UI 组件**: Element Plus
+- **CSS 框架**: TailwindCSS
+- **状态管理**: Pinia
+- **路由**: Vue Router 4
+
+### 开发工具
+
+- **包管理**: Go Modules, npm
+- **代码规范**: ESLint, Prettier
+- **版本控制**: Git
+
+---
+
+## 📝 常见问题
+
+### Q: Docker 容器如何访问宿主机的 Ollama？
+
+A: 使用 `http://host.docker.internal:11434/v1` 作为 Base URL。注意两点：
+
+1. 宿主机 Ollama 需监听 `0.0.0.0`：`export OLLAMA_HOST=0.0.0.0:11434 && ollama serve`
+2. Linux 用户使用 `docker run` 需添加：`--add-host=host.docker.internal:host-gateway`
+
+详见：[DOCKER_HOST_ACCESS.md](docs/DOCKER_HOST_ACCESS.md)
+
+### Q: FFmpeg 未安装或找不到？
+
+A: 确保 FFmpeg 已安装并在 PATH 环境变量中。运行 `ffmpeg -version` 验证。
+
+### Q: 前端无法连接后端 API？
+
+A: 检查后端是否启动，端口是否正确。开发模式下前端代理配置在 `web/vite.config.ts`。
+
+### Q: 数据库表未创建？
+
+A: GORM 会在首次启动时自动创建表，检查日志确认迁移是否成功。
+
+---
+
+## � 更新日志 / Changelog
+
+### v1.0.2 (2026-01-16)
+
+#### 🚀 重大更新
+
+- SQLite 纯 Go 驱动（`modernc.org/sqlite`），支持 `CGO_ENABLED=0` 跨平台编译
+- 优化并发性能（WAL 模式），解决 "database is locked" 错误
+- Docker 跨平台支持 `host.docker.internal` 访问宿主机服务
+- 精简文档和部署指南
+
+### v1.0.1 (2026-01-14)
+
+#### 🐛 Bug Fixes / 🔧 Improvements
+
+- 修复视频生成 API 响应解析问题
+- 添加 OpenAI Sora 视频端点配置
+- 优化错误处理和日志输出
+
+---
+
+## 🤝 贡献指南
+
+欢迎提交 Issue 和 Pull Request！
+
+1. Fork 本项目
+2. 创建特性分支 (`git checkout -b feature/AmazingFeature`)
+3. 提交改动 (`git commit -m 'Add some AmazingFeature'`)
+4. 推送到分支 (`git push origin feature/AmazingFeature`)
+5. 开启 Pull Request
+
+---
+
+## API 配置站点
+
+2 分钟完成配置：[API 聚合站点](https://api.chatfire.site/models)
+
+---
+
+## 👨‍💻 关于我们
+
+**AI 火宝 - AI 工作室创业中**
+
+- 🏠 **位置**: 中国南京
+- 🚀 **状态**: 创业中
+- 📧 **Email**: [18550175439@163.com](mailto:18550175439@163.com)
+- 💬 **微信**: dangbao1117 （私人微信不解答任何技术问题）
+- 🐙 **GitHub**: [https://github.com/chatfire-AI/huobao-drama](https://github.com/chatfire-AI/huobao-drama)
+
+> _"让 AI 帮我们做更有创造力的事"_
+
+## 项目交流群
+
+![项目交流群](drama.png)
+
+- 提交 [Issue](../../issues)
+- 发送邮件至项目维护者
+
+---
+
+<div align="center">
+
+**⭐ 如果这个项目对你有帮助，请给一个 Star！**
+
+## Star History
+
+[![Star History Chart](https://api.star-history.com/svg?repos=chatfire-AI/huobao-drama&type=date&legend=top-left)](https://www.star-history.com/#chatfire-AI/huobao-drama&type=date&legend=top-left)
+Made with ❤️ by Huobao Team
+
+</div>
diff --git a/README-EN.md b/README-EN.md
deleted file mode 100644
index a676b57..0000000
--- a/README-EN.md
+++ /dev/null
@@ -1,577 +0,0 @@
-# 🎬 Huobao Drama - AI Short Drama Production Platform
-
-<div align="center">
-
-**Full-stack AI Short Drama Automation Platform Based on Go + Vue3**
-
-[![Go Version](https://img.shields.io/badge/Go-1.23+-00ADD8?style=flat&logo=go)](https://golang.org)
-[![Vue Version](https://img.shields.io/badge/Vue-3.x-4FC08D?style=flat&logo=vue.js)](https://vuejs.org)
-[![License](https://img.shields.io/badge/License-CC%20BY--NC--SA%204.0-lightgrey.svg)](https://creativecommons.org/licenses/by-nc-sa/4.0/)
-
-[Features](#features) • [Quick Start](#quick-start) • [Deployment](#deployment)
-
-[简体中文](README.md) | [English](README-EN.md) | [日本語](README-JA.md)
-
-</div>
-
----
-
-## 📖 About
-
-Huobao Drama is an AI-powered short drama production platform that automates the entire workflow from script generation, character design, storyboarding to video composition.
-
-### 🎯 Core Features
-
-- **🤖 AI-Driven**: Parse scripts using large language models to extract characters, scenes, and storyboards
-- **🎨 Intelligent Creation**: AI-generated character portraits and scene backgrounds
-- **📹 Video Generation**: Automatic storyboard video generation using text-to-video and image-to-video models
-- **🔄 Complete Workflow**: End-to-end production workflow from idea to final video
-
-### 🛠️ Technical Architecture
-
-Based on **DDD (Domain-Driven Design)** with clear layering:
-
-```
-├── API Layer (Gin HTTP)
-├── Application Service Layer (Business Logic)
-├── Domain Layer (Domain Models)
-└── Infrastructure Layer (Database, External Services)
-```
-
-### 🎥 Demo Videos
-
-Experience AI short drama generation:
-
-<div align="center">
-
-**Sample Work 1**
-
-<video src="https://ffile.chatfire.site/cf/public/20260114094337396.mp4" controls width="640"></video>
-
-**Sample Work 2**
-
-<video src="https://ffile.chatfire.site/cf/public/fcede75e8aeafe22031dbf78f86285b8.mp4" controls width="640"></video>
-
-[Watch Video 1](https://ffile.chatfire.site/cf/public/20260114094337396.mp4) | [Watch Video 2](https://ffile.chatfire.site/cf/public/fcede75e8aeafe22031dbf78f86285b8.mp4)
-
-</div>
-
----
-
-## ✨ Features
-
-### 🎭 Character Management
-
-- ✅ AI-generated character portraits
-- ✅ Batch character generation
-- ✅ Character image upload and management
-
-### 🎬 Storyboard Production
-
-- ✅ Automatic storyboard script generation
-- ✅ Scene descriptions and shot design
-- ✅ Storyboard image generation (text-to-image)
-- ✅ Frame type selection (first frame/key frame/last frame/panel)
-
-### 🎥 Video Generation
-
-- ✅ Automatic image-to-video generation
-- ✅ Video composition and editing
-- ✅ Transition effects
-
-### 📦 Asset Management
-
-- ✅ Unified asset library management
-- ✅ Local storage support
-- ✅ Asset import/export
-- ✅ Task progress tracking
-
----
-
-## 🚀 Quick Start
-
-### 📋 Prerequisites
-
-| Software    | Version | Description                     |
-| ----------- | ------- | ------------------------------- |
-| **Go**      | 1.23+   | Backend runtime                 |
-| **Node.js** | 18+     | Frontend build environment      |
-| **npm**     | 9+      | Package manager                 |
-| **FFmpeg**  | 4.0+    | Video processing (**Required**) |
-| **SQLite**  | 3.x     | Database (built-in)             |
-
-#### Installing FFmpeg
-
-**macOS:**
-
-```bash
-brew install ffmpeg
-```
-
-**Ubuntu/Debian:**
-
-```bash
-sudo apt update
-sudo apt install ffmpeg
-```
-
-**Windows:**
-Download from [FFmpeg Official Site](https://ffmpeg.org/download.html) and configure environment variables
-
-Verify installation:
-
-```bash
-ffmpeg -version
-```
-
-### ⚙️ Configuration
-
-Copy and edit the configuration file:
-
-```bash
-cp configs/config.example.yaml configs/config.yaml
-vim configs/config.yaml
-```
-
-Configuration file format (`configs/config.yaml`):
-
-```yaml
-app:
-  name: "Huobao Drama API"
-  version: "1.0.0"
-  debug: true # Set to true for development, false for production
-
-server:
-  port: 5678
-  host: "0.0.0.0"
-  cors_origins:
-    - "http://localhost:3012"
-  read_timeout: 600
-  write_timeout: 600
-
-database:
-  type: "sqlite"
-  path: "./data/drama_generator.db"
-  max_idle: 10
-  max_open: 100
-
-storage:
-  type: "local"
-  local_path: "./data/storage"
-  base_url: "http://localhost:5678/static"
-
-ai:
-  default_text_provider: "openai"
-  default_image_provider: "openai"
-  default_video_provider: "doubao"
-```
-
-**Key Configuration Items:**
-
-- `app.debug`: Debug mode switch (recommended true for development)
-- `server.port`: Service port
-- `server.cors_origins`: Allowed CORS origins for frontend
-- `database.path`: SQLite database file path
-- `storage.local_path`: Local file storage path
-- `storage.base_url`: Static resource access URL
-- `ai.default_*_provider`: AI service provider configuration (API keys configured in Web UI)
-
-### 📥 Installation
-
-```bash
-# Clone the project
-git clone https://github.com/chatfire-AI/huobao-drama.git
-cd huobao-drama
-
-# Install Go dependencies
-go mod download
-
-# Install frontend dependencies
-cd web
-npm install
-cd ..
-```
-
-### 🎯 Starting the Project
-
-#### Method 1: Development Mode (Recommended)
-
-**Frontend and backend separation with hot reload**
-
-```bash
-# Terminal 1: Start backend service
-go run main.go
-
-# Terminal 2: Start frontend dev server
-cd web
-npm run dev
-```
-
-- Frontend: `http://localhost:3012`
-- Backend API: `http://localhost:5678/api/v1`
-- Frontend automatically proxies API requests to backend
-
-#### Method 2: Single Service Mode
-
-**Backend serves both API and frontend static files**
-
-```bash
-# 1. Build frontend
-cd web
-npm run build
-cd ..
-
-# 2. Start service
-go run main.go
-```
-
-Access: `http://localhost:5678`
-
-### 🗄️ Database Initialization
-
-Database tables are automatically created on first startup (using GORM AutoMigrate), no manual migration needed.
-
----
-
-## 📦 Deployment
-
-### 🐳 Docker Deployment (Recommended)
-
-#### Method 1: Docker Compose (Recommended)
-
-```bash
-# Start services
-docker-compose up -d
-
-# View logs
-docker-compose logs -f
-
-# Stop services
-docker-compose down
-```
-
-#### Method 2: Docker Command
-
-> **Note**: Linux users need to add `--add-host=host.docker.internal:host-gateway` to access host services
-
-```bash
-# Run from Docker Hub
-docker run -d \
-  --name huobao-drama \
-  -p 5678:5678 \
-  -v $(pwd)/data:/app/data \
-  --restart unless-stopped \
-  huobao/huobao-drama:latest
-
-# View logs
-docker logs -f huobao-drama
-```
-
-**Local Build** (optional):
-
-```bash
-docker build -t huobao-drama:latest .
-docker run -d --name huobao-drama -p 5678:5678 -v $(pwd)/data:/app/data huobao-drama:latest
-```
-
-**Docker Deployment Advantages:**
-
-- ✅ Ready to use with default configuration
-- ✅ Environment consistency, avoiding dependency issues
-- ✅ One-click start, no need to install Go, Node.js, FFmpeg
-- ✅ Easy to migrate and scale
-- ✅ Automatic health checks and restarts
-- ✅ Automatic file permission handling
-
-#### 🔗 Accessing Host Services (Ollama/Local Models)
-
-The container is configured to access host services using `http://host.docker.internal:PORT`.
-
-**Configuration Steps:**
-
-1. **Start service on host (listen on all interfaces)**
-
-   ```bash
-   export OLLAMA_HOST=0.0.0.0:11434 && ollama serve
-   ```
-
-2. **Frontend AI Service Configuration**
-   - Base URL: `http://host.docker.internal:11434/v1`
-   - Provider: `openai`
-   - Model: `qwen2.5:latest`
-
----
-
-### 🏭 Traditional Deployment
-
-#### 1. Build
-
-```bash
-# 1. Build frontend
-cd web
-npm run build
-cd ..
-
-# 2. Compile backend
-go build -o huobao-drama .
-```
-
-Generated files:
-
-- `huobao-drama` - Backend executable
-- `web/dist/` - Frontend static files (embedded in backend)
-
-#### 2. Prepare Deployment Files
-
-Files to upload to server:
-
-```
-huobao-drama            # Backend executable
-configs/config.yaml     # Configuration file
-data/                   # Data directory (optional, auto-created on first run)
-```
-
-#### 3. Server Configuration
-
-```bash
-# Upload files to server
-scp huobao-drama user@server:/opt/huobao-drama/
-scp configs/config.yaml user@server:/opt/huobao-drama/configs/
-
-# SSH to server
-ssh user@server
-
-# Modify configuration file
-cd /opt/huobao-drama
-vim configs/config.yaml
-# Set mode to production
-# Configure domain and storage path
-
-# Create data directory and set permissions (Important!)
-# Note: Replace YOUR_USER with actual user running the service (e.g., www-data, ubuntu, deploy)
-sudo mkdir -p /opt/huobao-drama/data/storage
-sudo chown -R YOUR_USER:YOUR_USER /opt/huobao-drama/data
-sudo chmod -R 755 /opt/huobao-drama/data
-
-# Grant execute permission
-chmod +x huobao-drama
-
-# Start service
-./huobao-drama
-```
-
-#### 4. Manage Service with systemd
-
-Create service file `/etc/systemd/system/huobao-drama.service`:
-
-```ini
-[Unit]
-Description=Huobao Drama Service
-After=network.target
-
-[Service]
-Type=simple
-User=YOUR_USER
-WorkingDirectory=/opt/huobao-drama
-ExecStart=/opt/huobao-drama/huobao-drama
-Restart=on-failure
-RestartSec=10
-
-# Environment variables (optional)
-# Environment="GIN_MODE=release"
-
-[Install]
-WantedBy=multi-user.target
-```
-
-Start service:
-
-```bash
-sudo systemctl daemon-reload
-sudo systemctl enable huobao-drama
-sudo systemctl start huobao-drama
-sudo systemctl status huobao-drama
-```
-
-**⚠️ Common Issue: SQLite Write Permission Error**
-
-If you encounter `attempt to write a readonly database` error:
-
-```bash
-# 1. Check current user running the service
-sudo systemctl status huobao-drama | grep "Main PID"
-ps aux | grep huobao-drama
-
-# 2. Fix permissions (replace YOUR_USER with actual username)
-sudo chown -R YOUR_USER:YOUR_USER /opt/huobao-drama/data
-sudo chmod -R 755 /opt/huobao-drama/data
-
-# 3. Verify permissions
-ls -la /opt/huobao-drama/data
-# Should show owner as the user running the service
-
-# 4. Restart service
-sudo systemctl restart huobao-drama
-```
-
-**Reason:**
-
-- SQLite requires write permission on both the database file **and** its directory
-- Needs to create temporary files in the directory (e.g., `-wal`, `-journal`)
-- **Key**: Ensure systemd `User` matches data directory owner
-
-**Common Usernames:**
-
-- Ubuntu/Debian: `www-data`, `ubuntu`
-- CentOS/RHEL: `nginx`, `apache`
-- Custom deployment: `deploy`, `app`, current logged-in user
-
-#### 5. Nginx Reverse Proxy
-
-```nginx
-server {
-    listen 80;
-    server_name your-domain.com;
-
-    location / {
-        proxy_pass http://localhost:5678;
-        proxy_set_header Host $host;
-        proxy_set_header X-Real-IP $remote_addr;
-        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
-    }
-
-    # Direct access to static files
-    location /static/ {
-        alias /opt/huobao-drama/data/storage/;
-    }
-}
-```
-
----
-
-## 🎨 Tech Stack
-
-### Backend
-
-- **Language**: Go 1.23+
-- **Web Framework**: Gin 1.9+
-- **ORM**: GORM
-- **Database**: SQLite
-- **Logging**: Zap
-- **Video Processing**: FFmpeg
-- **AI Services**: OpenAI, Gemini, Doubao, etc.
-
-### Frontend
-
-- **Framework**: Vue 3.4+
-- **Language**: TypeScript 5+
-- **Build Tool**: Vite 5
-- **UI Components**: Element Plus
-- **CSS Framework**: TailwindCSS
-- **State Management**: Pinia
-- **Router**: Vue Router 4
-
-### Development Tools
-
-- **Package Management**: Go Modules, npm
-- **Code Standards**: ESLint, Prettier
-- **Version Control**: Git
-
----
-
-## 📝 FAQ
-
-### Q: How can Docker containers access Ollama on the host?
-
-A: Use `http://host.docker.internal:11434/v1` as Base URL. Note two things:
-
-1. Host Ollama needs to listen on `0.0.0.0`: `export OLLAMA_HOST=0.0.0.0:11434 && ollama serve`
-2. Linux users using `docker run` need to add: `--add-host=host.docker.internal:host-gateway`
-
-See: [DOCKER_HOST_ACCESS.md](docs/DOCKER_HOST_ACCESS.md)
-
-### Q: FFmpeg not installed or not found?
-
-A: Ensure FFmpeg is installed and in the PATH environment variable. Verify with `ffmpeg -version`.
-
-### Q: Frontend cannot connect to backend API?
-
-A: Check if backend is running and port is correct. In development mode, frontend proxy config is in `web/vite.config.ts`.
-
-### Q: Database tables not created?
-
-A: GORM automatically creates tables on first startup, check logs to confirm migration success.
-
----
-
-## 📋 Changelog
-
-### v1.0.2 (2026-01-16)
-
-#### 🚀 Major Updates
-
-- Pure Go SQLite driver (`modernc.org/sqlite`), supports `CGO_ENABLED=0` cross-platform compilation
-- Optimized concurrency performance (WAL mode), resolved "database is locked" errors
-- Docker cross-platform support for `host.docker.internal` to access host services
-- Streamlined documentation and deployment guides
-
-### v1.0.1 (2026-01-14)
-
-#### 🐛 Bug Fixes / 🔧 Improvements
-
-- Fixed video generation API response parsing issues
-- Added OpenAI Sora video endpoint configuration
-- Optimized error handling and logging
-
----
-
-## 🤝 Contributing
-
-Issues and Pull Requests are welcome!
-
-1. Fork this project
-2. Create a feature branch (`git checkout -b feature/AmazingFeature`)
-3. Commit your changes (`git commit -m 'Add some AmazingFeature'`)
-4. Push to the branch (`git push origin feature/AmazingFeature`)
-5. Open a Pull Request
-
----
-
-## API Configuration Site
-
-Configure in 2 minutes: [API Aggregation Site](https://api.chatfire.site/models)
-
----
-
-## 👨‍💻 About Us
-
-**AI Huobao - AI Studio Startup**
-
-- 🏠 **Location**: Nanjing, China
-- 🚀 **Status**: Startup in Progress
-- 📧 **Email**: [18550175439@163.com](mailto:18550175439@163.com)
-- 💬 **WeChat**: dangbao1117 (Personal WeChat - No technical support)
-- 🐙 **GitHub**: [https://github.com/chatfire-AI/huobao-drama](https://github.com/chatfire-AI/huobao-drama)
-
-> _"Let AI help us do more creative things"_
-
-## Community Group
-
-![Community Group](drama.png)
-
-- Submit [Issue](../../issues)
-- Email project maintainers
-
----
-
-<div align="center">
-
-**⭐ If this project helps you, please give it a Star!**
-
-## Star History
-
-[![Star History Chart](https://api.star-history.com/svg?repos=chatfire-AI/huobao-drama&type=date&legend=top-left)](https://www.star-history.com/#chatfire-AI/huobao-drama&type=date&legend=top-left)
-
-Made with ❤️ by Huobao Team
-
-</div>
diff --git a/README-JA.md b/README-JA.md
index b9b1754..42e454b 100644
--- a/README-JA.md
+++ b/README-JA.md
@@ -10,7 +10,7 @@
 
 [機能](#機能) • [クイックスタート](#クイックスタート) • [デプロイ](#デプロイ)
 
-[简体中文](README.md) | [English](README-EN.md) | [日本語](README-JA.md)
+[简体中文](README-CN.md) | [English](README.md) | [日本語](README-JA.md)
 
 </div>
 
diff --git a/README.md b/README.md
index d8d745a..e62f8a1 100644
--- a/README.md
+++ b/README.md
@@ -1,108 +1,106 @@
-# 🎬 Huobao Drama - AI 短剧生成平台
+# 🎬 Huobao Drama - AI Short Drama Production Platform
 
 <div align="center">
 
-**基于 Go + Vue3 的全栈 AI 短剧自动化生产平台**
+**Full-stack AI Short Drama Automation Platform Based on Go + Vue3**
 
 [![Go Version](https://img.shields.io/badge/Go-1.23+-00ADD8?style=flat&logo=go)](https://golang.org)
 [![Vue Version](https://img.shields.io/badge/Vue-3.x-4FC08D?style=flat&logo=vue.js)](https://vuejs.org)
 [![License](https://img.shields.io/badge/License-CC%20BY--NC--SA%204.0-lightgrey.svg)](https://creativecommons.org/licenses/by-nc-sa/4.0/)
 
-[功能特性](#功能特性) • [快速开始](#快速开始) • [部署指南](#部署指南)
+[Features](#features) • [Quick Start](#quick-start) • [Deployment](#deployment)
 
-[简体中文](README.md) | [English](README-EN.md) | [日本語](README-JA.md)
+[简体中文](README-CN.md) | [English](README.md) | [日本語](README-JA.md)
 
 </div>
 
 ---
 
-## 📖 项目简介 / About
-
-Huobao Drama 是一个基于 AI 的短剧自动化生产平台，实现从剧本生成、角色设计、分镜制作到视频合成的全流程自动化。
+## 📖 About
 
 Huobao Drama is an AI-powered short drama production platform that automates the entire workflow from script generation, character design, storyboarding to video composition.
 
-### 🎯 核心价值 / Core Features
+### 🎯 Core Features
 
-- **🤖 AI 驱动 / AI-Driven**：使用大语言模型解析剧本，提取角色、场景和分镜信息 | Parse scripts using large language models to extract characters, scenes, and storyboards
-- **🎨 智能创作 / Intelligent Creation**：AI 绘图生成角色形象和场景背景 | AI-generated character portraits and scene backgrounds
-- **📹 视频生成 / Video Generation**：基于文生视频和图生视频模型自动生成分镜视频 | Automatic storyboard video generation using text-to-video and image-to-video models
-- **🔄 工作流 / Workflow**：完整的短剧制作工作流，从创意到成片一站式完成 | Complete production workflow from idea to final video
+- **🤖 AI-Driven**: Parse scripts using large language models to extract characters, scenes, and storyboards
+- **🎨 Intelligent Creation**: AI-generated character portraits and scene backgrounds
+- **📹 Video Generation**: Automatic storyboard video generation using text-to-video and image-to-video models
+- **🔄 Complete Workflow**: End-to-end production workflow from idea to final video
 
-### 🛠️ 技术架构
+### 🛠️ Technical Architecture
 
-采用**DDD 领域驱动设计**，清晰分层：
+Based on **DDD (Domain-Driven Design)** with clear layering:
 
 ```
-├── API层 (Gin HTTP)
-├── 应用服务层 (Business Logic)
-├── 领域层 (Domain Models)
-└── 基础设施层 (Database, External Services)
+├── API Layer (Gin HTTP)
+├── Application Service Layer (Business Logic)
+├── Domain Layer (Domain Models)
+└── Infrastructure Layer (Database, External Services)
 ```
 
-### 🎥 作品展示 / Demo Videos
+### 🎥 Demo Videos
 
-体验 AI 短剧生成效果：
+Experience AI short drama generation:
 
 <div align="center">
 
-**示例作品 1**
+**Sample Work 1**
 
 <video src="https://ffile.chatfire.site/cf/public/20260114094337396.mp4" controls width="640"></video>
 
-**示例作品 2**
+**Sample Work 2**
 
 <video src="https://ffile.chatfire.site/cf/public/fcede75e8aeafe22031dbf78f86285b8.mp4" controls width="640"></video>
 
-[点击观看视频 1](https://ffile.chatfire.site/cf/public/20260114094337396.mp4) | [点击观看视频 2](https://ffile.chatfire.site/cf/public/fcede75e8aeafe22031dbf78f86285b8.mp4)
+[Watch Video 1](https://ffile.chatfire.site/cf/public/20260114094337396.mp4) | [Watch Video 2](https://ffile.chatfire.site/cf/public/fcede75e8aeafe22031dbf78f86285b8.mp4)
 
 </div>
 
 ---
 
-## ✨ 功能特性
+## ✨ Features
 
-### 🎭 角色管理
+### 🎭 Character Management
 
-- ✅ AI 生成角色形象
-- ✅ 批量角色生成
-- ✅ 角色图片上传和管理
+- ✅ AI-generated character portraits
+- ✅ Batch character generation
+- ✅ Character image upload and management
 
-### 🎬 分镜制作
+### 🎬 Storyboard Production
 
-- ✅ 自动生成分镜脚本
-- ✅ 场景描述和镜头设计
-- ✅ 分镜图片生成（文生图）
-- ✅ 帧类型选择（首帧/关键帧/尾帧/分镜板）
+- ✅ Automatic storyboard script generation
+- ✅ Scene descriptions and shot design
+- ✅ Storyboard image generation (text-to-image)
+- ✅ Frame type selection (first frame/key frame/last frame/panel)
 
-### 🎥 视频生成
+### 🎥 Video Generation
 
-- ✅ 图生视频自动生成
-- ✅ 视频合成和剪辑
-- ✅ 转场效果
+- ✅ Automatic image-to-video generation
+- ✅ Video composition and editing
+- ✅ Transition effects
 
-### 📦 资源管理
+### 📦 Asset Management
 
-- ✅ 素材库统一管理
-- ✅ 本地存储支持
-- ✅ 资源导入导出
-- ✅ 任务进度追踪
+- ✅ Unified asset library management
+- ✅ Local storage support
+- ✅ Asset import/export
+- ✅ Task progress tracking
 
 ---
 
-## 🚀 快速开始
+## 🚀 Quick Start
 
-### 📋 环境要求
+### 📋 Prerequisites
 
-| 软件        | 版本要求 | 说明                 |
-| ----------- | -------- | -------------------- |
-| **Go**      | 1.23+    | 后端运行环境         |
-| **Node.js** | 18+      | 前端构建环境         |
-| **npm**     | 9+       | 包管理工具           |
-| **FFmpeg**  | 4.0+     | 视频处理（**必需**） |
-| **SQLite**  | 3.x      | 数据库（已内置）     |
+| Software    | Version | Description                     |
+| ----------- | ------- | ------------------------------- |
+| **Go**      | 1.23+   | Backend runtime                 |
+| **Node.js** | 18+     | Frontend build environment      |
+| **npm**     | 9+      | Package manager                 |
+| **FFmpeg**  | 4.0+    | Video processing (**Required**) |
+| **SQLite**  | 3.x     | Database (built-in)             |
 
-#### 安装 FFmpeg
+#### Installing FFmpeg
 
 **macOS:**
 
@@ -118,30 +116,30 @@ sudo apt install ffmpeg
 ```
 
 **Windows:**
-从 [FFmpeg 官网](https://ffmpeg.org/download.html) 下载并配置环境变量
+Download from [FFmpeg Official Site](https://ffmpeg.org/download.html) and configure environment variables
 
-验证安装：
+Verify installation:
 
 ```bash
 ffmpeg -version
 ```
 
-### ⚙️ 配置文件
+### ⚙️ Configuration
 
-复制并编辑配置文件：
+Copy and edit the configuration file:
 
 ```bash
 cp configs/config.example.yaml configs/config.yaml
 vim configs/config.yaml
 ```
 
-配置文件格式（`configs/config.yaml`）：
+Configuration file format (`configs/config.yaml`):
 
 ```yaml
 app:
   name: "Huobao Drama API"
   version: "1.0.0"
-  debug: true # 开发环境设为true，生产环境设为false
+  debug: true # Set to true for development, false for production
 
 server:
   port: 5678
@@ -168,143 +166,96 @@ ai:
   default_video_provider: "doubao"
 ```
 
-**重要配置项：**
+**Key Configuration Items:**
 
-- `app.debug`: 调试模式开关（开发环境建议设为 true）
-- `server.port`: 服务运行端口
-- `server.cors_origins`: 允许跨域访问的前端地址
-- `database.path`: SQLite 数据库文件路径
-- `storage.local_path`: 本地文件存储路径
-- `storage.base_url`: 静态资源访问 URL
-- `ai.default_*_provider`: AI 服务提供商配置（在 Web 界面中配置具体的 API Key）
+- `app.debug`: Debug mode switch (recommended true for development)
+- `server.port`: Service port
+- `server.cors_origins`: Allowed CORS origins for frontend
+- `database.path`: SQLite database file path
+- `storage.local_path`: Local file storage path
+- `storage.base_url`: Static resource access URL
+- `ai.default_*_provider`: AI service provider configuration (API keys configured in Web UI)
 
-### 📥 安装依赖
+### 📥 Installation
 
 ```bash
-# 克隆项目
+# Clone the project
 git clone https://github.com/chatfire-AI/huobao-drama.git
 cd huobao-drama
 
-# 安装Go依赖
+# Install Go dependencies
 go mod download
 
-# 安装前端依赖
+# Install frontend dependencies
 cd web
 npm install
 cd ..
 ```
 
-### 🎯 启动项目
+### 🎯 Starting the Project
 
-#### 方式一：开发模式（推荐）
+#### Method 1: Development Mode (Recommended)
 
-**前后端分离，支持热重载**
+**Frontend and backend separation with hot reload**
 
 ```bash
-# 终端1：启动后端服务
+# Terminal 1: Start backend service
 go run main.go
 
-# 终端2：启动前端开发服务器
+# Terminal 2: Start frontend dev server
 cd web
 npm run dev
 ```
 
-- 前端地址: `http://localhost:3012`
-- 后端 API: `http://localhost:5678/api/v1`
-- 前端自动代理 API 请求到后端
+- Frontend: `http://localhost:3012`
+- Backend API: `http://localhost:5678/api/v1`
+- Frontend automatically proxies API requests to backend
 
-#### 方式二：单服务模式
+#### Method 2: Single Service Mode
 
-**后端同时提供 API 和前端静态文件**
+**Backend serves both API and frontend static files**
 
 ```bash
-# 1. 构建前端
+# 1. Build frontend
 cd web
 npm run build
 cd ..
 
-# 2. 启动服务
+# 2. Start service
 go run main.go
 ```
 
-访问: `http://localhost:5678`
+Access: `http://localhost:5678`
 
-### 🗄️ 数据库初始化
+### 🗄️ Database Initialization
 
-数据库表会在首次启动时自动创建（使用 GORM AutoMigrate），无需手动迁移。
+Database tables are automatically created on first startup (using GORM AutoMigrate), no manual migration needed.
 
 ---
 
-## 📦 部署指南
+## 📦 Deployment
 
-### 🐳 Docker 部署（推荐）
+### 🐳 Docker Deployment (Recommended)
 
-#### 🚀 国内网络加速（可选）
-
-如果您在国内网络环境下，Docker 拉取镜像和安装依赖可能较慢。可以通过配置镜像源加速构建过程。
-
-**步骤1：创建环境变量文件**
+#### Method 1: Docker Compose (Recommended)
 
 ```bash
-cp .env.example .env
-```
-
-**步骤2：编辑 `.env` 文件，取消注释需要的镜像源**
-
-```bash
-# 启用 Docker Hub 镜像（推荐）
-DOCKER_REGISTRY=docker.1ms.run/
-
-# 启用 npm 镜像
-NPM_REGISTRY=https://registry.npmmirror.com/
-
-# 启用 Go 代理
-GO_PROXY=https://goproxy.cn,direct
-
-# 启用 Alpine 镜像
-ALPINE_MIRROR=mirrors.aliyun.com
-```
-
-**步骤3：使用 docker compose 构建（必须）**
-
-```bash
-docker compose build
-```
-
-> **重要说明**：
-> - ⚠️ 必须使用 `docker compose build` 才能自动加载 `.env` 文件中的镜像源配置
-> - ❌ 如果使用 `docker build` 命令，需要手动传递 `--build-arg` 参数
-> - ✅ 推荐始终使用 `docker compose build` 进行构建
-
-**效果对比**：
-
-| 操作 | 不配置镜像源 | 配置镜像源后 |
-|------|------------|-------------|
-| 拉取基础镜像 | 5-30分钟 | 1-5分钟 |
-| 安装 npm 依赖 | 可能失败 | 快速成功 |
-| 下载 Go 依赖 | 5-10分钟 | 30秒-1分钟 |
-
-> **注意**：国外用户请勿配置镜像源，使用默认配置即可。
-
-#### 方式一：Docker Compose（推荐）
-
-```bash
-# 启动服务
+# Start services
 docker-compose up -d
 
-# 查看日志
+# View logs
 docker-compose logs -f
 
-# 停止服务
+# Stop services
 docker-compose down
 ```
 
-#### 方式二：Docker 命令
+#### Method 2: Docker Command
 
-> **注意**：Linux 用户需添加 `--add-host=host.docker.internal:host-gateway` 以访问宿主机服务
+> **Note**: Linux users need to add `--add-host=host.docker.internal:host-gateway` to access host services
 
 ```bash
-# 从 Docker Hub 运行
+# Run from Docker Hub
 docker run -d \
   --name huobao-drama \
   -p 5678:5678 \
@@ -312,106 +263,106 @@ docker run -d \
   --restart unless-stopped \
   huobao/huobao-drama:latest
 
-# 查看日志
+# View logs
 docker logs -f huobao-drama
 ```
 
-**本地构建**（可选）：
+**Local Build** (optional):
 
 ```bash
 docker build -t huobao-drama:latest .
 docker run -d --name huobao-drama -p 5678:5678 -v $(pwd)/data:/app/data huobao-drama:latest
 ```
 
-**Docker 部署优势：**
+**Docker Deployment Advantages:**
 
-- ✅ 开箱即用，内置默认配置
-- ✅ 环境一致性，避免依赖问题
-- ✅ 一键启动，无需安装 Go、Node.js、FFmpeg
-- ✅ 易于迁移和扩展
-- ✅ 自动健康检查和重启
-- ✅ 自动处理文件权限，无需手动配置
+- ✅ Ready to use with default configuration
+- ✅ Environment consistency, avoiding dependency issues
+- ✅ One-click start, no need to install Go, Node.js, FFmpeg
+- ✅ Easy to migrate and scale
+- ✅ Automatic health checks and restarts
+- ✅ Automatic file permission handling
 
-#### 🔗 访问宿主机服务（Ollama/本地模型）
+#### 🔗 Accessing Host Services (Ollama/Local Models)
 
-容器已配置支持访问宿主机服务，直接使用 `http://host.docker.internal:端口号` 即可。
+The container is configured to access host services using `http://host.docker.internal:PORT`.
 
-**配置步骤：**
+**Configuration Steps:**
 
-1. **宿主机启动服务（监听所有接口）**
+1. **Start service on host (listen on all interfaces)**
 
    ```bash
    export OLLAMA_HOST=0.0.0.0:11434 && ollama serve
    ```
 
-2. **前端 AI 服务配置**
+2. **Frontend AI Service Configuration**
    - Base URL: `http://host.docker.internal:11434/v1`
    - Provider: `openai`
    - Model: `qwen2.5:latest`
 
 ---
 
-### 🏭 传统部署方式
+### 🏭 Traditional Deployment
 
-#### 1. 编译构建
+#### 1. Build
 
 ```bash
-# 1. 构建前端
+# 1. Build frontend
 cd web
 npm run build
 cd ..
 
-# 2. 编译后端
+# 2. Compile backend
 go build -o huobao-drama .
 ```
 
-生成文件：
+Generated files:
 
-- `huobao-drama` - 后端可执行文件
-- `web/dist/` - 前端静态文件（已嵌入后端）
+- `huobao-drama` - Backend executable
+- `web/dist/` - Frontend static files (embedded in backend)
 
-#### 2. 准备部署文件
+#### 2. Prepare Deployment Files
 
-需要上传到服务器的文件：
+Files to upload to server:
 
 ```
-huobao-drama            # 后端可执行文件
-configs/config.yaml     # 配置文件
-data/                   # 数据目录（可选，首次运行自动创建）
+huobao-drama            # Backend executable
+configs/config.yaml     # Configuration file
+data/                   # Data directory (optional, auto-created on first run)
 ```
 
-#### 3. 服务器配置
+#### 3. Server Configuration
 
 ```bash
-# 上传文件到服务器
+# Upload files to server
 scp huobao-drama user@server:/opt/huobao-drama/
 scp configs/config.yaml user@server:/opt/huobao-drama/configs/
 
-# SSH登录服务器
+# SSH to server
 ssh user@server
 
-# 修改配置文件
+# Modify configuration file
 cd /opt/huobao-drama
 vim configs/config.yaml
-# 设置mode为production
-# 配置域名和存储路径
+# Set mode to production
+# Configure domain and storage path
 
-# 创建数据目录并设置权限（重要！）
-# 注意：将 YOUR_USER 替换为实际运行服务的用户名（如 www-data、ubuntu、deploy 等）
+# Create data directory and set permissions (Important!)
+# Note: Replace YOUR_USER with actual user running the service (e.g., www-data, ubuntu, deploy)
 sudo mkdir -p /opt/huobao-drama/data/storage
 sudo chown -R YOUR_USER:YOUR_USER /opt/huobao-drama/data
 sudo chmod -R 755 /opt/huobao-drama/data
 
-# 赋予执行权限
+# Grant execute permission
 chmod +x huobao-drama
 
-# 启动服务
+# Start service
 ./huobao-drama
 ```
 
-#### 4. 使用 systemd 管理服务
+#### 4. Manage Service with systemd
 
-创建服务文件 `/etc/systemd/system/huobao-drama.service`:
+Create service file `/etc/systemd/system/huobao-drama.service`:
 
 ```ini
 [Unit]
@@ -426,14 +377,14 @@ ExecStart=/opt/huobao-drama/huobao-drama
 Restart=on-failure
 RestartSec=10
 
-# 环境变量（可选）
+# Environment variables (optional)
 # Environment="GIN_MODE=release"
 
 [Install]
 WantedBy=multi-user.target
 ```
 
-启动服务：
+Start service:
 
 ```bash
 sudo systemctl daemon-reload
@@ -442,40 +393,40 @@ sudo systemctl start huobao-drama
 sudo systemctl status huobao-drama
 ```
 
-**⚠️ 常见问题：SQLite 写权限错误**
+**⚠️ Common Issue: SQLite Write Permission Error**
 
-如果遇到 `attempt to write a readonly database` 错误：
+If you encounter `attempt to write a readonly database` error:
 
 ```bash
-# 1. 确认当前运行服务的用户
+# 1. Check current user running the service
 sudo systemctl status huobao-drama | grep "Main PID"
 ps aux | grep huobao-drama
 
-# 2. 修复权限（将 YOUR_USER 替换为实际用户名）
+# 2. Fix permissions (replace YOUR_USER with actual username)
 sudo chown -R YOUR_USER:YOUR_USER /opt/huobao-drama/data
 sudo chmod -R 755 /opt/huobao-drama/data
 
-# 3. 验证权限
+# 3. Verify permissions
 ls -la /opt/huobao-drama/data
-# 应该显示所有者为运行服务的用户
+# Should show owner as the user running the service
 
-# 4. 重启服务
+# 4. Restart service
 sudo systemctl restart huobao-drama
 ```
 
-**原因说明**：
+**Reason:**
 
-- SQLite 需要对数据库文件 **和** 所在目录都有写权限
-- 需要在目录中创建临时文件（如 `-wal`、`-journal`）
-- **关键**：确保 systemd 配置中的 `User` 与数据目录所有者一致
+- SQLite requires write permission on both the database file **and** its directory
+- Needs to create temporary files in the directory (e.g., `-wal`, `-journal`)
+- **Key**: Ensure systemd `User` matches data directory owner
 
-**常用用户名**：
+**Common Usernames:**
 
-- Ubuntu/Debian: `www-data`、`ubuntu`
-- CentOS/RHEL: `nginx`、`apache`
-- 自定义部署: `deploy`、`app`、当前登录用户
+- Ubuntu/Debian: `www-data`, `ubuntu`
+- CentOS/RHEL: `nginx`, `apache`
+- Custom deployment: `deploy`, `app`, current logged-in user
 
-#### 5. Nginx 反向代理
+#### 5. Nginx Reverse Proxy
 
 ```nginx
 server {
@@ -489,7 +440,7 @@ server {
         proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
     }
 
-    # 静态文件直接访问
+    # Direct access to static files
     location /static/ {
         alias /opt/huobao-drama/data/storage/;
     }
@@ -498,128 +449,129 @@ server {
 
 ---
 
-## 🎨 技术栈
+## 🎨 Tech Stack
 
-### 后端技术
+### Backend
 
-- **语言**: Go 1.23+
-- **Web 框架**: Gin 1.9+
+- **Language**: Go 1.23+
+- **Web Framework**: Gin 1.9+
 - **ORM**: GORM
-- **数据库**: SQLite
-- **日志**: Zap
-- **视频处理**: FFmpeg
-- **AI 服务**: OpenAI、Gemini、火山等
+- **Database**: SQLite
+- **Logging**: Zap
+- **Video Processing**: FFmpeg
+- **AI Services**: OpenAI, Gemini, Doubao, etc.
 
-### 前端技术
+### Frontend
 
-- **框架**: Vue 3.4+
-- **语言**: TypeScript 5+
-- **构建工具**: Vite 5
-- **UI 组件**: Element Plus
-- **CSS 框架**: TailwindCSS
-- **状态管理**: Pinia
-- **路由**: Vue Router 4
+- **Framework**: Vue 3.4+
+- **Language**: TypeScript 5+
+- **Build Tool**: Vite 5
+- **UI Components**: Element Plus
+- **CSS Framework**: TailwindCSS
+- **State Management**: Pinia
+- **Router**: Vue Router 4
 
-### 开发工具
+### Development Tools
 
-- **包管理**: Go Modules, npm
-- **代码规范**: ESLint, Prettier
-- **版本控制**: Git
+- **Package Management**: Go Modules, npm
+- **Code Standards**: ESLint, Prettier
+- **Version Control**: Git
 
 ---
 
-## 📝 常见问题
+## 📝 FAQ
 
-### Q: Docker 容器如何访问宿主机的 Ollama？
+### Q: How can Docker containers access Ollama on the host?
 
-A: 使用 `http://host.docker.internal:11434/v1` 作为 Base URL。注意两点：
+A: Use `http://host.docker.internal:11434/v1` as Base URL. Note two things:
 
-1. 宿主机 Ollama 需监听 `0.0.0.0`：`export OLLAMA_HOST=0.0.0.0:11434 && ollama serve`
-2. Linux 用户使用 `docker run` 需添加：`--add-host=host.docker.internal:host-gateway`
+1. Host Ollama needs to listen on `0.0.0.0`: `export OLLAMA_HOST=0.0.0.0:11434 && ollama serve`
+2. Linux users using `docker run` need to add: `--add-host=host.docker.internal:host-gateway`
 
-详见：[DOCKER_HOST_ACCESS.md](docs/DOCKER_HOST_ACCESS.md)
+See: [DOCKER_HOST_ACCESS.md](docs/DOCKER_HOST_ACCESS.md)
 
-### Q: FFmpeg 未安装或找不到？
+### Q: FFmpeg not installed or not found?
 
-A: 确保 FFmpeg 已安装并在 PATH 环境变量中。运行 `ffmpeg -version` 验证。
+A: Ensure FFmpeg is installed and in the PATH environment variable. Verify with `ffmpeg -version`.
 
-### Q: 前端无法连接后端 API？
+### Q: Frontend cannot connect to backend API?
 
-A: 检查后端是否启动，端口是否正确。开发模式下前端代理配置在 `web/vite.config.ts`。
+A: Check if backend is running and port is correct. In development mode, frontend proxy config is in `web/vite.config.ts`.
 
-### Q: 数据库表未创建？
+### Q: Database tables not created?
 
-A: GORM 会在首次启动时自动创建表，检查日志确认迁移是否成功。
+A: GORM automatically creates tables on first startup, check logs to confirm migration success.
 
 ---
 
-## � 更新日志 / Changelog
+## 📋 Changelog
 
 ### v1.0.2 (2026-01-16)
 
-#### 🚀 重大更新
+#### 🚀 Major Updates
 
-- SQLite 纯 Go 驱动（`modernc.org/sqlite`），支持 `CGO_ENABLED=0` 跨平台编译
-- 优化并发性能（WAL 模式），解决 "database is locked" 错误
-- Docker 跨平台支持 `host.docker.internal` 访问宿主机服务
-- 精简文档和部署指南
+- Pure Go SQLite driver (`modernc.org/sqlite`), supports `CGO_ENABLED=0` cross-platform compilation
+- Optimized concurrency performance (WAL mode), resolved "database is locked" errors
+- Docker cross-platform support for `host.docker.internal` to access host services
+- Streamlined documentation and deployment guides
 
 ### v1.0.1 (2026-01-14)
 
 #### 🐛 Bug Fixes / 🔧 Improvements
 
-- 修复视频生成 API 响应解析问题
-- 添加 OpenAI Sora 视频端点配置
-- 优化错误处理和日志输出
+- Fixed video generation API response parsing issues
+- Added OpenAI Sora video endpoint configuration
+- Optimized error handling and logging
 
 ---
 
-## 🤝 贡献指南
+## 🤝 Contributing
 
-欢迎提交 Issue 和 Pull Request！
+Issues and Pull Requests are welcome!
 
-1. Fork 本项目
-2. 创建特性分支 (`git checkout -b feature/AmazingFeature`)
-3. 提交改动 (`git commit -m 'Add some AmazingFeature'`)
-4. 推送到分支 (`git push origin feature/AmazingFeature`)
-5. 开启 Pull Request
+1. Fork this project
+2. Create a feature branch (`git checkout -b feature/AmazingFeature`)
+3. Commit your changes (`git commit -m 'Add some AmazingFeature'`)
+4. Push to the branch (`git push origin feature/AmazingFeature`)
+5. Open a Pull Request
 
 ---
 
-## API 配置站点
+## API Configuration Site
 
-2 分钟完成配置：[API 聚合站点](https://api.chatfire.site/models)
+Configure in 2 minutes: [API Aggregation Site](https://api.chatfire.site/models)
 
 ---
 
-## 👨‍💻 关于我们
+## 👨‍💻 About Us
 
-**AI 火宝 - AI 工作室创业中**
+**AI Huobao - AI Studio Startup**
 
-- 🏠 **位置**: 中国南京
-- 🚀 **状态**: 创业中
+- 🏠 **Location**: Nanjing, China
+- 🚀 **Status**: Startup in Progress
 - 📧 **Email**: [18550175439@163.com](mailto:18550175439@163.com)
-- 💬 **微信**: dangbao1117 （私人微信不解答任何技术问题）
+- 💬 **WeChat**: dangbao1117 (Personal WeChat - No technical support)
 - 🐙 **GitHub**: [https://github.com/chatfire-AI/huobao-drama](https://github.com/chatfire-AI/huobao-drama)
 
-> _"让 AI 帮我们做更有创造力的事"_
+> _"Let AI help us do more creative things"_
 
-## 项目交流群
+## Community Group
 
-![项目交流群](drama.png)
+![Community Group](drama.png)
 
-- 提交 [Issue](../../issues)
-- 发送邮件至项目维护者
+- Submit [Issue](../../issues)
+- Email project maintainers
 
 ---
 
 <div align="center">
 
-**⭐ 如果这个项目对你有帮助，请给一个 Star！**
+**⭐ If this project helps you, please give it a Star!**
 
 ## Star History
 
 [![Star History Chart](https://api.star-history.com/svg?repos=chatfire-AI/huobao-drama&type=date&legend=top-left)](https://www.star-history.com/#chatfire-AI/huobao-drama&type=date&legend=top-left)
+
 Made with ❤️ by Huobao Team
 
 </div>
diff --git a/api/handlers/scene.go b/api/handlers/scene.go
index 1367992..cf9fa4f 100644
--- a/api/handlers/scene.go
+++ b/api/handlers/scene.go
@@ -74,6 +74,28 @@ func (h *SceneHandler) GenerateSceneImage(c *gin.Context) {
 	})
 }
 
+func (h *SceneHandler) UpdateScenePrompt(c *gin.Context) {
+	sceneID := c.Param("scene_id")
+
+	var req services2.UpdateScenePromptRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		response.BadRequest(c, "Invalid request")
+		return
+	}
+
+	if err := h.sceneService.UpdateScenePrompt(sceneID, &req); err != nil {
+		h.log.Errorw("Failed to update scene prompt", "error", err, "scene_id", sceneID)
+		if err.Error() == "scene not found" {
+			response.NotFound(c, "场景不存在")
+			return
+		}
+		response.InternalError(c, err.Error())
+		return
+	}
+
+	response.Success(c, gin.H{"message": "场景提示词已更新"})
+}
+
 func (h *SceneHandler) DeleteScene(c *gin.Context) {
 	sceneID := c.Param("scene_id")
 
diff --git a/api/routes/routes.go b/api/routes/routes.go
index 295e7d4..a97fc92 100644
--- a/api/routes/routes.go
+++ b/api/routes/routes.go
@@ -137,6 +137,7 @@ func SetupRouter(cfg *config.Config, db *gorm.DB, log *logger.Logger, localStora
 		scenes := api.Group("/scenes")
 		{
 			scenes.PUT("/:scene_id", sceneHandler.UpdateScene)
+			scenes.PUT("/:scene_id/prompt", sceneHandler.UpdateScenePrompt)
 			scenes.DELETE("/:scene_id", sceneHandler.DeleteScene)
 			scenes.POST("/generate-image", sceneHandler.GenerateSceneImage)
 		}
diff --git a/application/services/storyboard_composition_service.go b/application/services/storyboard_composition_service.go
index 0b203c3..9264729 100644
--- a/application/services/storyboard_composition_service.go
+++ b/application/services/storyboard_composition_service.go
@@ -399,6 +399,28 @@ func (s *StoryboardCompositionService) GenerateSceneImage(req *GenerateSceneImag
 	return nil, fmt.Errorf("image generation service not available")
 }
 
+type UpdateScenePromptRequest struct {
+	Prompt string `json:"prompt"`
+}
+
+func (s *StoryboardCompositionService) UpdateScenePrompt(sceneID string, req *UpdateScenePromptRequest) error {
+	var scene models.Scene
+	if err := s.db.Where("id = ?", sceneID).First(&scene).Error; err != nil {
+		if err == gorm.ErrRecordNotFound {
+			return fmt.Errorf("scene not found")
+		}
+		return fmt.Errorf("failed to find scene: %w", err)
+	}
+
+	scene.Prompt = req.Prompt
+	if err := s.db.Save(&scene).Error; err != nil {
+		return fmt.Errorf("failed to update scene prompt: %w", err)
+	}
+
+	s.log.Infow("Scene prompt updated", "scene_id", sceneID, "prompt", req.Prompt)
+	return nil
+}
+
 func (s *StoryboardCompositionService) DeleteScene(sceneID string) error {
 	var scene models.Scene
 	if err := s.db.Where("id = ?", sceneID).First(&scene).Error; err != nil {
diff --git a/pkg/image/gemini_image_client.go b/pkg/image/gemini_image_client.go
index 3ffa427..08e9926 100644
--- a/pkg/image/gemini_image_client.go
+++ b/pkg/image/gemini_image_client.go
@@ -107,7 +107,7 @@ func NewGeminiImageClient(baseURL, apiKey, model, endpoint string) *GeminiImageC
 
 func (c *GeminiImageClient) GenerateImage(prompt string, opts ...ImageOption) (*ImageResult, error) {
 	options := &ImageOptions{
-		Size:    "1024x1024",
+		Size:    "1920x1920",
 		Quality: "standard",
 	}
 
diff --git a/pkg/image/volcengine_image_client.go b/pkg/image/volcengine_image_client.go
index 87b920d..9373516 100644
--- a/pkg/image/volcengine_image_client.go
+++ b/pkg/image/volcengine_image_client.go
@@ -63,7 +63,7 @@ func NewVolcEngineImageClient(baseURL, apiKey, model, endpoint, queryEndpoint st
 
 func (c *VolcEngineImageClient) GenerateImage(prompt string, opts ...ImageOption) (*ImageResult, error) {
 	options := &ImageOptions{
-		Size:    "1024x1024",
+		Size:    "1920x1920",
 		Quality: "standard",
 	}
 
diff --git a/web/src/api/drama.ts b/web/src/api/drama.ts
index 34a3acf..12d0cfa 100644
--- a/web/src/api/drama.ts
+++ b/web/src/api/drama.ts
@@ -108,8 +108,12 @@ export const dramaAPI = {
     return request.put(`/scenes/${sceneId}`, data)
   },
 
-  generateSceneImage(data: { scene_id: string; prompt?: string; model?: string }) {
-    return request.post('/scenes/generate-image', data)
+  generateSceneImage(data: { scene_id: number; prompt?: string; model?: string }) {
+    return request.post<{ image_generation: { id: number } }>('/scenes/generate-image', data)
+  },
+
+  updateScenePrompt(sceneId: string, prompt: string) {
+    return request.put(`/scenes/${sceneId}/prompt`, { prompt })
   },
 
   deleteScene(sceneId: string) {
diff --git a/web/src/locales/zh-CN.ts b/web/src/locales/zh-CN.ts
index 8eeaa5f..bd8bd14 100644
--- a/web/src/locales/zh-CN.ts
+++ b/web/src/locales/zh-CN.ts
@@ -68,7 +68,9 @@ export default {
     tip: '提示',
     status: '状态',
     createdAt: '创建时间',
-    updatedAt: '更新时间'
+    updatedAt: '更新时间',
+    name: '名称',
+    description: '描述'
   },
   settings: {
     title: '设置',
diff --git a/web/src/views/drama/DramaManagement.vue b/web/src/views/drama/DramaManagement.vue
index 8bf61d7..41fcd43 100644
--- a/web/src/views/drama/DramaManagement.vue
+++ b/web/src/views/drama/DramaManagement.vue
@@ -241,10 +241,10 @@
     <el-dialog v-model="addSceneDialogVisible" :title="$t('common.add')" width="600px">
       <el-form :model="newScene" label-width="100px">
         <el-form-item :label="$t('common.name')">
-          <el-input v-model="newScene.name" :placeholder="$t('common.name')" />
+          <el-input v-model="newScene.location" :placeholder="$t('common.name')" />
         </el-form-item>
         <el-form-item :label="$t('common.description')">
-          <el-input v-model="newScene.description" type="textarea" :rows="4" :placeholder="$t('common.description')" />
+          <el-input v-model="newScene.prompt" type="textarea" :rows="4" :placeholder="$t('common.description')" />
         </el-form-item>
       </el-form>
       <template #footer>
@@ -285,8 +285,8 @@ const newCharacter = ref({
 })
 
 const newScene = ref({
-  name: '',
-  description: ''
+  location: '',
+  prompt: ''
 })
 
 const episodesCount = computed(() => drama.value?.episodes?.length || 0)
@@ -494,14 +494,14 @@ const deleteCharacter = async (character: any) => {
 
 const openAddSceneDialog = () => {
   newScene.value = {
-    name: '',
-    description: ''
+    location: '',
+    prompt: ''
   }
   addSceneDialogVisible.value = true
 }
 
 const addScene = async () => {
-  if (!newScene.value.name.trim()) {
+  if (!newScene.value.location.trim()) {
     ElMessage.warning('请输入场景名称')
     return
   }
@@ -517,7 +517,11 @@ const addScene = async () => {
 }
 
 const editScene = (scene: any) => {
-  ElMessage.info('编辑功能开发中')
+  newScene.value = {
+    location: scene.location || scene.name || '',
+    prompt: scene.prompt || scene.description || ''
+  }
+  addSceneDialogVisible.value = true
 }
 
 const deleteScene = async (scene: any) => {
diff --git a/web/src/views/drama/EpisodeWorkflow.vue b/web/src/views/drama/EpisodeWorkflow.vue
index bd8c7d7..8ab6f9b 100644
--- a/web/src/views/drama/EpisodeWorkflow.vue
+++ b/web/src/views/drama/EpisodeWorkflow.vue
@@ -1601,8 +1601,9 @@ const saveShotEdit = async () => {
 // 对话框相关方法
 const openPromptDialog = (item: any, type: 'character' | 'scene') => {
   currentEditItem.value = item
+  currentEditItem.value.name = item.name || item.location
   currentEditType.value = type
-  editPrompt.value = item.appearance || item.description || ''
+  editPrompt.value = item.prompt || item.appearance || item.description || ''
   promptDialogVisible.value = true
 }
 
@@ -1614,10 +1615,32 @@ const savePrompt = async () => {
       })
       await generateCharacterImage(currentEditItem.value.id)
     } else {
-      await dramaAPI.generateSceneImage({ 
+      // 1. 先保存场景提示词
+      await dramaAPI.updateScenePrompt(currentEditItem.value.id.toString(), editPrompt.value)
+      
+      // 2. 生成场景图片
+      const model = selectedImageModel.value || undefined
+      const response = await dramaAPI.generateSceneImage({ 
         scene_id: parseInt(currentEditItem.value.id),
-        prompt: editPrompt.value
+        prompt: editPrompt.value,
+        model
       })
+      const imageGenId = response.image_generation?.id
+      
+      // 3. 轮询图片生成状态
+      if (imageGenId) {
+        ElMessage.info('场景图片生成中，请稍候...')
+        generatingSceneImages.value[currentEditItem.value.id] = true
+        pollImageStatus(imageGenId, async () => {
+          await loadDramaData()
+          ElMessage.success('场景图片生成完成！')
+        }).finally(() => {
+          generatingSceneImages.value[currentEditItem.value.id] = false
+        })
+      } else {
+        ElMessage.success('场景图片生成已启动')
+        await loadDramaData()
+      }
     }
     promptDialogVisible.value = false
   } catch (error: any) {
diff --git a/web/src/views/drama/ProfessionalEditor.vue b/web/src/views/drama/ProfessionalEditor.vue
index 4ccd6bc..65cd5f4 100644
--- a/web/src/views/drama/ProfessionalEditor.vue
+++ b/web/src/views/drama/ProfessionalEditor.vue
@@ -360,6 +360,32 @@
                     <!-- 首帧 -->
                     <div v-show="selectedVideoFrameType === 'first'" class="image-scroll-container"
                       style="max-height: 280px; overflow-y: auto; overflow-x: hidden;">
+                      
+                      <!-- 上一镜头尾帧推荐（紧凑版） -->
+                      <div v-if="previousStoryboardLastFrames.length > 0" class="previous-frame-section">
+                        <div style="display: flex; align-items: center; gap: 6px; margin-bottom: 6px;">
+                          <el-tag size="small" type="primary">
+                            上一镜头 #{{ previousStoryboard?.storyboard_number }} 尾帧
+                          </el-tag>
+                          <span class="hint-text">点击添加为首帧参考</span>
+                        </div>
+                        <div style="display: flex; gap: 8px; flex-wrap: wrap;">
+                          <div v-for="img in previousStoryboardLastFrames" :key="'prev-' + img.id" 
+                            class="reference-item"
+                            :class="{ selected: selectedImagesForVideo.includes(img.id) }"
+                            style="position: relative; border: 2px solid #1890ff; border-radius: 4px; overflow: hidden; cursor: pointer;"
+                            @click="selectPreviousLastFrame(img)">
+                            <el-image :src="img.image_url" fit="cover"
+                              style="width: 60px; height: 40px; display: block; pointer-events: none;" />
+                            <div v-if="selectedImagesForVideo.includes(img.id)" 
+                              style="position: absolute; top: 0; right: 0; background: #52c41a; color: #fff; font-size: 10px; padding: 1px 4px;">
+                              ✓
+                            </div>
+                          </div>
+                        </div>
+                      </div>
+                      
+                      <!-- 当前镜头首帧列表 -->
                       <div class="reference-grid"
                         style="display: grid; grid-template-columns: repeat(4, 1fr); gap: 12px; max-width: 600px;">
                         <div
@@ -378,7 +404,7 @@
                         </div>
                       </div>
                       <el-empty
-                        v-if="!videoReferenceImages.some(i => i.status === 'completed' && i.image_url && i.frame_type === 'first')"
+                        v-if="!videoReferenceImages.some(i => i.status === 'completed' && i.image_url && i.frame_type === 'first') && previousStoryboardLastFrames.length === 0"
                         description="暂无首帧图片" size="small" />
                     </div>
 
@@ -883,7 +909,7 @@ import { ElMessage, ElMessageBox } from 'element-plus'
 import {
   ArrowLeft, Plus, Picture, VideoPlay, VideoPause, View, Setting,
   Upload, MagicStick, VideoCamera, ZoomIn, ZoomOut, Top, Bottom, Check, Close, Right,
-  Timer, Calendar, Clock, Loading, WarningFilled, Delete
+  Timer, Calendar, Clock, Loading, WarningFilled, Delete, Connection
 } from '@element-plus/icons-vue'
 import { dramaAPI } from '@/api/drama'
 import { generateFramePrompt, type FrameType } from '@/api/frame'
@@ -1226,6 +1252,67 @@ const currentStoryboard = computed(() => {
   return storyboards.value.find(s => String(s.id) === String(currentStoryboardId.value)) || null
 })
 
+// 获取上一个镜头
+const previousStoryboard = computed(() => {
+  if (!currentStoryboardId.value || storyboards.value.length < 2) return null
+  const currentIndex = storyboards.value.findIndex(s => String(s.id) === String(currentStoryboardId.value))
+  if (currentIndex <= 0) return null
+  return storyboards.value[currentIndex - 1]
+})
+
+// 上一个镜头的尾帧图片列表（支持多个）
+const previousStoryboardLastFrames = ref<any[]>([])
+
+// 加载上一个镜头的尾帧
+const loadPreviousStoryboardLastFrame = async () => {
+  if (!previousStoryboard.value) {
+    previousStoryboardLastFrames.value = []
+    return
+  }
+  try {
+    const result = await imageAPI.listImages({
+      storyboard_id: previousStoryboard.value.id,
+      frame_type: 'last',
+      page: 1,
+      page_size: 10
+    })
+    const images = result.items || []
+    previousStoryboardLastFrames.value = images.filter((img: any) => img.status === 'completed' && img.image_url)
+  } catch (error) {
+    console.error('加载上一镜头尾帧失败:', error)
+    previousStoryboardLastFrames.value = []
+  }
+}
+
+// 选择上一镜头尾帧作为首帧参考
+const selectPreviousLastFrame = (img: any) => {
+  // 检查是否已选中，已选中则取消
+  const currentIndex = selectedImagesForVideo.value.indexOf(img.id)
+  if (currentIndex > -1) {
+    selectedImagesForVideo.value.splice(currentIndex, 1)
+    ElMessage.success('已取消首帧参考')
+    return
+  }
+
+  // 参考handleImageSelect的逻辑，根据模式处理
+  if (!selectedReferenceMode.value || selectedReferenceMode.value === 'single') {
+    // 单图模式或未选模式：直接替换
+    selectedImagesForVideo.value = [img.id]
+  } else if (selectedReferenceMode.value === 'first_last') {
+    // 首尾帧模式：作为首帧参考
+    selectedImagesForVideo.value = [img.id]
+  } else if (selectedReferenceMode.value === 'multiple') {
+    // 多图模式：添加到列表
+    const capability = currentModelCapability.value
+    if (capability && selectedImagesForVideo.value.length >= capability.maxImages) {
+      ElMessage.warning(`最多只能选择${capability.maxImages}张图片`)
+      return
+    }
+    selectedImagesForVideo.value.push(img.id)
+  }
+  ElMessage.success('已添加为首帧参考')
+}
+
 // 监听帧类型切换，从存储中加载或清空
 watch(selectedFrameType, (newType) => {
   // 切换帧类型时，停止之前的轮询，避免旧结果覆盖新帧类型
@@ -1269,6 +1356,7 @@ watch(currentStoryboard, async (newStoryboard) => {
     generatedImages.value = []
     generatedVideos.value = []
     videoReferenceImages.value = []
+    previousStoryboardLastFrames.value = []
     return
   }
 
@@ -1297,6 +1385,9 @@ watch(currentStoryboard, async (newStoryboard) => {
 
   // 加载该分镜的视频列表
   await loadStoryboardVideos(newStoryboard.id)
+
+  // 加载上一镜头的尾帧
+  await loadPreviousStoryboardLastFrame()
 })
 
 // 监听提示词变化，自动保存到sessionStorage
@@ -1801,13 +1892,17 @@ const selectedImageObjects = computed(() => {
 const firstFrameSlotImage = computed(() => {
   if (selectedImagesForVideo.value.length === 0) return null
   const firstImageId = selectedImagesForVideo.value[0]
-  return videoReferenceImages.value.find(img => img.id === firstImageId)
+  // 同时搜索当前镜头图片和上一镜头尾帧
+  return videoReferenceImages.value.find(img => img.id === firstImageId) 
+    || previousStoryboardLastFrames.value.find(img => img.id === firstImageId)
 })
 
 // 首尾帧模式：获取尾帧图片
 const lastFrameSlotImage = computed(() => {
   if (!selectedLastImageForVideo.value) return null
+  // 同时搜索当前镜头图片和上一镜头尾帧
   return videoReferenceImages.value.find(img => img.id === selectedLastImageForVideo.value)
+    || previousStoryboardLastFrames.value.find(img => img.id === selectedLastImageForVideo.value)
 })
 
 // 移除已选择的图片
@@ -1846,7 +1941,9 @@ const generateVideo = async () => {
   // 获取第一张选中的图片（仅在需要图片的模式下）
   let selectedImage = null
   if (selectedReferenceMode.value !== 'none' && selectedImagesForVideo.value.length > 0) {
+    // 同时搜索当前镜头图片和上一镜头尾帧
     selectedImage = videoReferenceImages.value.find(img => img.id === selectedImagesForVideo.value[0])
+      || previousStoryboardLastFrames.value.find(img => img.id === selectedImagesForVideo.value[0])
     if (!selectedImage || !selectedImage.image_url) {
       ElMessage.error('请选择有效的参考图片')
       return
@@ -1878,9 +1975,11 @@ const generateVideo = async () => {
         break
 
       case 'first_last':
-        // 首尾帧模式
+        // 首尾帧模式（同时搜索当前镜头图片和上一镜头尾帧）
         const firstImage = videoReferenceImages.value.find(img => img.id === selectedImagesForVideo.value[0])
+          || previousStoryboardLastFrames.value.find(img => img.id === selectedImagesForVideo.value[0])
         const lastImage = videoReferenceImages.value.find(img => img.id === selectedLastImageForVideo.value)
+          || previousStoryboardLastFrames.value.find(img => img.id === selectedLastImageForVideo.value)
 
         if (firstImage?.image_url) {
           requestParams.first_frame_url = firstImage.image_url
@@ -3947,6 +4046,19 @@ onBeforeUnmount(() => {
       }
     }
 
+    .previous-frame-section {
+      margin-bottom: 12px;
+      padding: 8px;
+      background: var(--bg-secondary);
+      border: 1px solid var(--border-primary);
+      border-radius: 6px;
+
+      .hint-text {
+        color: var(--text-muted);
+        font-size: 11px;
+      }
+    }
+
     .reference-grid {
       display: grid !important;
       grid-template-columns: repeat(4, 1fr) !important;