v4.12.0 - 项目清理和模板优化

更新内容

🧹 项目清理

• 删除过期的文档和临时文件
• 清理不再使用的配置文件和资源

📝 模板更新

• 更新 Pod2Post 模板配置
• 优化模板文档结构

🔧 代码优化

• 改进项目结构和代码组织
• 提升代码可维护性

📊 统计

• 删除 39 个文件
• 移除 10,615 行代码
• 新增 1,079 行代码

🤖 Generated with Claude Code

v4.3.0 - Custom Template Async API

🎉 主要新功能

自定义模板异步API

• 新增 /api/generate/custom/async 接口
• 支持ZIP文件上传和模板处理
• 两阶段生成流程：
  1. 第一阶段：生成HTML文件
  2. 第二阶段：自动将图片转换为base64嵌入

路径占位符支持

• [user] - 用户card工作目录
• [文件名] - ZIP中的文件路径
• "用户card路径" - 中文占位符支持
• [用户card路径] - 方括号中文占位符

🔧 技术改进

• Session管理优化：使用共享session保持上下文连续性
• 中文支持：完整支持中文文件名和路径的ZIP解压
• 日志优化：减少冗余日志输出，提升可读性
• 环境变量：添加默认配置，简化部署
• 命令执行：分离命令和回车发送，提升稳定性

📝 文档与测试

• 完整的API设计文档：terminal-backend/docs/custom-template-api-design.md
• 测试脚本：terminal-backend/test/test-custom-api.js
• Mock prompt支持快速测试

🐛 修复

• 修复session提前销毁问题
• 修复node-pty的offData方法调用错误
• 优化文件检测逻辑

📦 依赖更新

• 新增 yauzl - ZIP解压库
• 新增 yazl - ZIP创建库

---

Full Changelog: v4.2.0...v4.3.0

v4.2.0 - 音视频转录服务 & 移动端优化

🎉 新功能

音视频转录服务

• 🎙️ 集成阿里云 SenseVoice 模型，支持音视频转文字
• 📁 支持多种格式：mp3, mp4, wav, m4a, aac, opus, flac, ogg, amr, webm, mov, avi, mkv, wmv, flv
• 🔄 批量转录功能，支持同时处理多个文件
• 📊 任务队列管理系统，支持异步任务处理
• 🔍 任务状态实时查询和统计信息
• 🔁 失败任务重试机制

🐛 修复和优化

移动端体验优化

• 📱 修复移动端分享链接无法打开的问题
• 🔗 移动端改为当前页面跳转（window.location.href），更可靠
• 🚀 PC端保持新窗口打开，失败时自动降级

生产环境准备

• 🗑️ 移除 MobileDebugPanel 调试组件
• 📋 确保 template-registry.jsonl 为模板配置唯一数据源
• 🎯 优化用户提示信息，准确反映实际行为

📝 技术改进

• 重构转录服务为模块化架构（SenseVoice 目录）
• 实现完整的任务生命周期管理
• 改进错误处理和日志记录

v1.0.0: Audio/Video Transcription Service

🎉 Release v1.0.0: Audio/Video Transcription Service

✨ Features

• 🎯 Complete transcription service - Full audio/video to text conversion
• 💾 OSS Integration - Automatic file upload to Aliyun OSS
• 🎙️ SenseVoice API - High-accuracy speech recognition
• ⏱️ Async Processing - Background task execution with progress tracking
• 📦 Batch Support - Process multiple files simultaneously
• 📝 Rich Output - Timestamped Markdown and SRT subtitle files

🛠️ Technical Components

• OSS Service: Handles file storage and secure URL generation
• SenseVoice Service: Manages transcription tasks and results
• Task Manager: Tracks async jobs and maintains task history
• REST API: Complete set of transcription endpoints

📋 Supported Formats

Audio: WAV, MP3, M4A, AAC, OPUS, FLAC, OGG, AMR
Video: MP4, MOV, AVI, MKV, WMV, FLV, WebM

📚 Documentation

• OSS Service Guide
• SenseVoice Service Guide
• Transcription API Guide

🚀 Quick Start

# Clone repository
git clone https://github.com/aixier/AI_Terminal.git
cd AI_Terminal/terminal-backend

# Install dependencies
npm install

# Configure environment
cp .env.example .env
# Edit .env with your API keys

# Start service
npm run dev

🔧 Configuration Required

• Aliyun SenseVoice API Key
• Aliyun OSS Access Keys
• OSS Bucket configuration

📈 What's Next

This release establishes the foundation for the AI Terminal transcription module. Next phase will focus on:

• Frontend integration
• Real-time streaming transcription
• Multi-language support expansion
• Performance optimizations

🙏 Acknowledgments

Built with Aliyun Cloud Services (OSS & SenseVoice)

---

Full Changelog: https://github.com/aixier/AI_Terminal/commits/v1.0.0-transcription

Full Changelog: v4.2.0...v1.0.0-transcription

v3.32: 社交分享功能完整集成

🎉 新功能

社交媒体分享功能

• 在 HtmlMessageCard 添加分享按钮
• 创建 SocialShareDialog 社交平台选择对话框
• 实现紧凑型 4 列网格布局设计
• 集成小红书分享完整流程

优化改进

• 统一分享逻辑与作品集页面保持一致
• 优化移动端响应式布局体验
• 简化对话框交互设计
• 支持平台热门标记和锁定状态显示

技术细节

• 使用 useXiaohongshuShare 组合式函数
• 集成 ShareDialog 分享结果编辑组件
• 确保分享参数和接口调用一致性

Full Changelog: v3.31...v3.32

v3.62.2 - CardPlanet Sandra JSON 增强

🚀 新功能

API 增强

• 🎯 为 cardplanet-Sandra-json 模板添加 pageinfo 字段
  • 在 /api/generate/card 响应中新增独立的 JSON 数据返回
  • HTML 内容保持在 content 字段
  • JSON 数据单独返回在 pageinfo 字段
  • 支持双文件输出模板的完整数据访问

优化改进

• 📊 调整文件检测逻辑，确保正确识别主文件类型
• 🔧 HTML 文件作为主文件，JSON 文件用于数据补充

📝 更新说明

此版本专注于改进 cardplanet-Sandra-json 模板的数据返回结构，使前端能够更方便地访问和处理生成的卡片数据。

🐛 修复

• 修复 HtmlContentViewer 组件模板语法错误（v3.62.1）

---

Docker 镜像: ai-terminal:v3.62.2

Full Changelog: v3.62.1...v3.62.2

v3.10.42 - 移动端文件预览增强版

🚀 AI Terminal v3.10.42 - 移动端文件预览增强版

✨ 主要更新

🔧 移动端预览功能全面升级

• 加载状态指示器: 为JSON/HTML文件预览添加进度条和加载状态
• 智能错误处理: 提供详细的解析错误提示，帮助用户快速定位问题
• 预览验证机制: 只有解析成功后才触发预览，避免无效预览
• 递归文件查找: 修复子文件夹嵌套结构下的文件查找问题

🎨 UI/UX 优化

• 选中状态增强: 移动端卡片选中采用绿色主题高亮，视觉反馈更明显
• 模板选择优化: 改用背景色变化替代不明显的勾号显示
• 统一交互体验: 所有文件类型（JSON、HTML、Markdown）预览交互一致

🐛 问题修复

• 修复移动端选中文件后点击"预览"提示"请先选择一个文件"的问题
• 修复嵌套文件夹结构下无法正确找到文件的问题
• 优化文件类型识别和格式验证逻辑

📱 移动端体验提升

预览功能

• 进度指示: 实时显示文件加载进度（0-100%）
• 错误分类: 针对不同错误类型提供具体解决建议
• 加载动画: 优雅的加载状态动画和进度条

选中反馈

• 绿色主题: 采用现代化的绿色渐变高亮效果
• 动态效果: 选中时的缩放和阴影动画
• 视觉层次: 清晰的选中状态标识

🛠️ 技术改进

代码优化

• 添加递归文件夹查找算法
• 增强异步加载和错误处理机制
• 优化文件内容验证逻辑
• 添加详细调试日志

性能提升

• 优化移动端预览加载速度
• 减少不必要的内容重新加载
• 改进错误恢复机制

🔄 兼容性

• 保持向后兼容性
• 支持所有主流移动设备
• 优化触摸交互体验

🚢 部署说明

Docker 部署

```bash
docker pull ai-terminal:v3.10.42
docker run -d --name ai-terminal-v3.10.42 -p 6000:6000 ai-terminal:v3.10.42
```

更新建议

建议从v3.10.41及以下版本更新，享受更流畅的移动端预览体验。

---

完整变更日志: v3.10.41...v3.10.42

v3.9.8 - Enhanced Connection Management & Mobile Terminal Optimization

🚀 AI Terminal v3.9.8 - Enhanced Connection Management

🆕 Major Features

🔌 Real-time Terminal Connection Management

• Visual Connection Status: Real-time WebSocket connection indicators with color-coded status
• Auto-Recovery System: Intelligent reconnection with exponential backoff (up to 10 attempts)
• Manual Recovery Controls: One-click reconnection and cursor refresh buttons
• Connection Monitoring: Automatic status checks every 2 seconds with user feedback

📱 Mobile Terminal Optimization

• Mobile-First Design: Comprehensive mobile device detection and optimization
• Touch-Optimized Controls: Responsive buttons and interface elements for mobile devices
• Virtual Keyboard Handling: Smart layout adjustments for mobile virtual keyboards
• Forced Cursor Visibility: CSS overrides to ensure cursor visibility on mobile devices
• Mobile Initialization: Dedicated 📱 button for mobile terminal reinitialization

🛡️ Docker Security & Deployment Strategy

• Dual-Image Architecture: Separated public and production Docker images for enhanced security
• Public Image: coopotfan/ai-terminal:latest - Clean image without embedded secrets
• Production Image: coopotfan/ai-terminal:production - Pre-configured with API credentials
• Security Enhancement: DockerfileProduct added to .gitignore to prevent credential exposure

🔧 Technical Improvements

XTerm.js Terminal Enhancements

• Enhanced Auto-Resize: Debounced resize handling with mobile orientation change support
• Improved Cursor Management: ANSI control sequences for reliable cursor display
• Better Focus Handling: Enhanced terminal focus and blur event management
• Session Persistence: Improved WebSocket session management with reconnection support

Frontend Improvements

• Connection Status Bar: Visual indicators for connected/disconnected/reconnecting states
• Keyboard Shortcuts: Added Ctrl+Shift+R for cursor refresh and terminal focus
• Mobile Responsiveness: Optimized layouts for screens from 320px to desktop
• Loading States: Better visual feedback during initialization and reconnection

Backend Stability

• Socket.IO Optimization: Forced polling transport for better cloud environment compatibility
• Health Check API: New /api/terminal/health endpoint for connection validation
• Error Recovery: Enhanced error handling and user guidance messages
• Connection Timeout: Configurable timeout settings with graceful degradation

🚀 Quick Start

Public image (development/testing):
docker run -d -p 6000:6000 -e ANTHROPIC_AUTH_TOKEN="your_token" coopotfan/ai-terminal:latest

Production image (ready-to-use):
docker run -d -p 6000:6000 coopotfan/ai-terminal:production

Full Changelog: v3.381...v3.9.8

🌟 v2.5 - Platform Vision: Universal CLI-to-API Platform

🌟 AI Terminal Platform Vision - Democratizing ALL Professional CLI Tools

🎯 Major Announcement: From Tool to Platform

AI Terminal is evolving! We're not just another Claude Code wrapper - we're building the universal platform for transforming ANY professional CLI tool into accessible APIs.

🚀 The Vision

The Problem

Professional CLI tools (Claude Code, Gemini CLI, Cursor, etc.) are incredibly powerful but:

• ❌ Require technical expertise to use
• ❌ Can't be integrated into applications
• ❌ Have high learning curves
• ❌ Work in isolation

Our Solution

One platform to API-fy them all!

Professional CLI Tools  →  AI Terminal Platform  →  REST/WebSocket APIs

📊 Platform Roadmap

CLI Tool	Status	Timeline	What You Get
Claude Code	✅ LIVE NOW	Available	Full API, Streaming, Production Ready
Gemini CLI	🚧 Development	Q1 2025	Multimodal AI, Vision API
Cursor CLI	📋 Planned	Q1 2025	Code Editing, Refactoring APIs
GPT CLI	📋 Planned	Q2 2025	GPT-4, Custom GPTs
Ollama	📋 Planned	Q2 2025	Local Models, Privacy First
Your Favorite CLI	💡	Request It!	Community Driven

✨ What's New in v2.5

📚 Platform Documentation

• VISION.md - Complete platform vision and strategy
• ROADMAP_2025.md - Detailed 2025 execution plan
• Universal CLI Adapter Architecture - How we'll support ANY CLI

🔥 Current Claude Code API Features

Still the best way to use Claude Code as an API:

• ✅ REST API with full streaming support
• ✅ 30-second Docker deployment
• ✅ Production-ready with session management
• ✅ Complete documentation and examples

🎨 Use Cases Unlocked

For Developers

// Use multiple AI tools in one workflow
const pipeline = await aiTerminal
  .use('claude-code').for('research')
  .use('gemini').for('visual-generation')
  .use('cursor').for('code-refactoring')
  .execute();

For Content Creators

No coding required - visual workflow builder coming Q2 2025!

For Enterprises

• Integrate ANY CLI tool into your workflow
• No vendor lock-in
• Self-hosted, secure, compliant

🚀 2025 Milestones

Q1 2025: Multi-AI Support

• Gemini CLI integration
• Cursor CLI integration
• Universal adapter framework
• 4+ CLI tools supported

Q2 2025: Accessibility

• No-code workflow builder
• Mobile applications
• Browser extensions
• 10+ CLI tools supported

Q3 2025: Enterprise

• Multi-tenant architecture
• SSO/SAML support
• Compliance certifications
• 20+ CLI tools supported

Q4 2025: Ecosystem

• Plugin marketplace
• Community adapters
• 100+ integrations
• Self-sustaining ecosystem

💡 Why This Matters

We're not just building another tool - we're democratizing access to professional software.

Imagine:

• Product managers using Cursor CLI without knowing commands
• Content creators accessing Gemini's multimodal AI through simple APIs
• Students learning with professional tools through friendly interfaces
• Enterprises integrating ANY CLI tool without custom development

🤝 Join the Revolution

For Developers

• ⭐ Star the project - help us reach 10,000 stars!
• 🔧 Build adapters for your favorite CLI tools
• 📝 Share your use cases
• 🐛 Report issues and contribute code

For Organizations

• 🏢 Join our pilot program
• 💡 Share your CLI tool requirements
• 🤝 Partner with us
• 💰 Sponsor development

Community First

This is an open-source, community-driven project. The core platform will ALWAYS be free and open source.

📈 Current Stats

• 🌟 Growing GitHub community
• 🚀 Claude Code API already in production use
• 💪 Battle-tested architecture
• 📚 Comprehensive documentation

🎯 The Goal

By end of 2025: The de facto standard for CLI tool API transformation

• 10,000+ developers using the platform
• 100+ enterprise customers
• 50+ CLI tools supported
• Thriving plugin ecosystem

🙏 Thank You!

To our early adopters and contributors - you're not just users, you're pioneers in democratizing professional tools!

Special thanks to everyone who starred, contributed, and spread the word. This is just the beginning!

📚 Resources

• 📖 Platform Vision Document
• 🗺️ 2025 Roadmap
• 🔧 Universal CLI Adapter Architecture
• 🚀 Claude Code API Quick Start

🚀 Get Started Today!

While we build the future, Claude Code API is ready NOW:

# Deploy in 30 seconds
docker run -d -p 8082:6000 aixier/ai-terminal:latest

# Test immediately
curl -X POST http://localhost:8082/api/generate/card \
  -H "Content-Type: application/json" \
  -d '{"topic": "Hello World"}'

---

Together, we're democratizing professional CLI tools, one API at a time.

🌟 Star us: https://github.com/aixier/AI_Terminal
💬 Join Discord: https://discord.gg/ai-terminal
🐦 Follow: @aiterminal

#AITerminal #ClaudeCodeAPI #CLItoAPI #OpenSource #DemocratizingTools

🔥 v2.4 - Claude Code API Release

🔥 Claude Code API - The Game Changer!

Transform Claude Code CLI into REST API

We're excited to announce Claude Code API - the first open-source solution to transform Anthropic's Claude Code CLI into powerful REST APIs!

🎯 Why This Matters

Claude Code is amazing, but it's limited to CLI. Now you can:

• Use Claude Code from any application via REST API
• Stream responses in real-time with Server-Sent Events
• Integrate into your existing workflow without CLI limitations
• Deploy in production with Docker in 30 seconds

✨ Key Features

🚀 Quick Start (30 seconds!)

# Deploy Claude Code API
docker run -d -p 8082:6000 aixier/ai-terminal:latest

# Test it immediately
curl -X POST http://localhost:8082/api/generate/card \
  -H "Content-Type: application/json" \
  -d '{"topic": "Your Topic Here"}'

🌊 Streaming Support

Real-time streaming with Server-Sent Events:

const eventSource = new EventSource('/api/generate/card/stream');
eventSource.addEventListener('output', (e) => {
    console.log('Claude:', JSON.parse(e.data));
});

📦 Non-Streaming Support

Traditional REST API for batch processing:

import requests
response = requests.post('http://localhost:8082/api/generate/card',
                         json={'topic': 'Python Best Practices'})
print(response.json())

📚 Complete Documentation

• 📖 Claude Code API Documentation
• 🚀 5-Minute Quick Start Guide
• 💻 Code Examples (Node.js/Python)

🎊 What's New in v2.4

Major Features

• Claude Code API - Complete REST API implementation
• Streaming & Non-Streaming - Both modes fully supported
• Auto Session Management - Handles Claude initialization automatically
• Production Ready - Battle-tested with proper error handling
• Complete Examples - Node.js and Python client examples

Improvements

• Better Claude initialization handling for both configured/unconfigured states
• Optimized parallel file detection for faster responses
• Enhanced documentation with SEO optimization
• Added 19 GitHub topics for better discoverability

💡 Use Cases

1. Content Generation Platforms

   • Generate blog posts, documentation, tutorials
   • Stream content in real-time to users

2. Developer Tools Integration

   • Add Claude Code to your IDE/editor
   • Create coding assistants

3. Educational Platforms

   • Interactive coding tutorials
   • Real-time code explanation

4. CI/CD Pipelines

   • Automated code review
   • Documentation generation

🔥 Why Choose AI Terminal for Claude Code API?

Feature	AI Terminal	DIY Solution	Direct CLI
REST API	✅ Built-in	🔧 Build yourself	❌
Streaming	✅ SSE Ready	🔧 Complex	❌
Docker	✅ One command	🔧 Setup needed	❌
Production	✅ Ready	❓ Depends	❌

📈 Stats

• 🚀 30-second deployment with Docker
• ⚡ 2-second first response time
• 🔄 100+ concurrent sessions supported
• 📦 Zero configuration required

🙏 Thank You!

This release wouldn't be possible without our amazing community. Special thanks to all contributors!

🚀 Get Started Now

# It's this simple\!
docker run -d -p 8082:6000 aixier/ai-terminal:latest

Then visit our Quick Start Guide to integrate Claude Code API into your application in 5 minutes!

📢 Spread the Word!

If you find this useful, please:

• ⭐ Star the repository
• 🐦 Tweet with #ClaudeCodeAPI
• 📝 Write about your use case
• 💬 Join our Discord

---

Keywords: Claude Code API, Claude API, Anthropic Claude, REST API, Streaming API, AI API, Claude Integration

Full Changelog: v2.3...v2.4