diff --git a/.github/PULL_REQUEST_TEMPLATE.md b/.github/PULL_REQUEST_TEMPLATE.md index 3c4c9080324..27755fc32b7 100644 --- a/.github/PULL_REQUEST_TEMPLATE.md +++ b/.github/PULL_REQUEST_TEMPLATE.md @@ -1,3 +1,4 @@ +$$ #### πŸ’» ε˜ζ›΄η±»εž‹ | Change Type @@ -26,3 +27,5 @@ Thank you for your Pull Request. Please provide a description above. 请添加与歀 Pull Request η›Έε…³ηš„θ‘₯充俑息 Add any other context about the Pull Request here. --> + +$$ \ No newline at end of file diff --git a/CLAUDE.md b/CLAUDE.md new file mode 100644 index 00000000000..b2db1de8c77 --- /dev/null +++ b/CLAUDE.md @@ -0,0 +1,879 @@ +# CLAUDE.md - NextChat Codebase Guide for AI Assistants + +> **Last Updated**: 2025-12-25 +> **Repository**: NextChat (formerly ChatGPT-Next-Web) +> **Version**: 2.16.1 + +## Table of Contents + +1. [Project Overview](#project-overview) +2. [Architecture & Technology Stack](#architecture--technology-stack) +3. [Project Structure](#project-structure) +4. [Development Workflows](#development-workflows) +5. [Key Conventions](#key-conventions) +6. [Common Tasks & Patterns](#common-tasks--patterns) +7. [Testing Guidelines](#testing-guidelines) +8. [Deployment](#deployment) +9. [AI Assistant Guidelines](#ai-assistant-guidelines) +10. [Important Files Reference](#important-files-reference) + +--- + +## Project Overview + +**NextChat** is a cross-platform AI chat application supporting 15+ AI providers including OpenAI, Anthropic (Claude), Google Gemini, DeepSeek, and others. It's designed as a privacy-first, lightweight ChatGPT web client with desktop application support via Tauri. + +### Key Features +- Multi-provider AI support (OpenAI, Claude, Gemini, DeepSeek, etc.) +- Privacy-first: all data stored locally in browser +- Desktop apps for Linux/Windows/macOS via Tauri +- PWA with offline capability +- Model Context Protocol (MCP) support +- Real-time chat with WebRTC audio +- Artifacts system for code/content preview +- 21 language translations +- Prompt templates (Masks) + +### Tech Stack Summary +- **Frontend**: Next.js 14.1.1 (App Router), React 18.2.0, TypeScript 5.2.2 +- **Styling**: SASS/SCSS with CSS Modules +- **State**: Zustand 4.3.8 +- **Desktop**: Tauri 1.5.11 (Rust) +- **Testing**: Jest 29.7.0 + React Testing Library + +--- + +## Architecture & Technology Stack + +### Frontend Architecture + +``` +Multi-Provider Pattern +β”œβ”€β”€ Dynamic API routing: /app/api/[provider]/[...path]/route.ts +β”œβ”€β”€ Platform implementations: /app/client/platforms/ +β”œβ”€β”€ Unified client API: getClientApi() +└── Supports 15+ providers +``` + +### State Management (Zustand Stores) + +Located in `/app/store/`: +- **chat.ts**: Chat sessions, messages, conversation history +- **config.ts**: App configuration and model settings +- **access.ts**: API keys and access control +- **mask.ts**: Prompt templates +- **plugin.ts**: Plugin management +- **sync.ts**: Cloud sync configuration + +### Component Architecture + +- **Dynamic imports** for code splitting (lazy-loaded routes) +- **CSS Modules** for component-scoped styling (`*.module.scss`) +- **ErrorBoundary** wrapping for fault isolation +- **Client-side routing** with React Router + +### API Design + +- **Edge Runtime** for fast API responses +- **Streaming support** for real-time chat responses +- **CORS headers** configured for cross-origin requests +- **Proxy rewrites** for various AI provider endpoints + +--- + +## Project Structure + +``` +/home/user/Mr.advokatio-/ +β”œβ”€β”€ app/ # Next.js app directory (main source) +β”‚ β”œβ”€β”€ api/ # API routes +β”‚ β”‚ β”œβ”€β”€ [provider]/ # Dynamic provider routing +β”‚ β”‚ β”œβ”€β”€ openai.ts # OpenAI API handler +β”‚ β”‚ β”œβ”€β”€ anthropic.ts # Anthropic/Claude API handler +β”‚ β”‚ └── config/ # Config endpoint +β”‚ β”œβ”€β”€ client/ # API client implementations +β”‚ β”‚ └── platforms/ # Individual provider clients +β”‚ β”‚ β”œβ”€β”€ openai.ts +β”‚ β”‚ β”œβ”€β”€ anthropic.ts +β”‚ β”‚ β”œβ”€β”€ google.ts +β”‚ β”‚ └── ... +β”‚ β”œβ”€β”€ components/ # React components +β”‚ β”‚ β”œβ”€β”€ home.tsx # Main app shell +β”‚ β”‚ β”œβ”€β”€ chat.tsx # Chat interface +β”‚ β”‚ β”œβ”€β”€ settings.tsx # Settings panel +β”‚ β”‚ β”œβ”€β”€ mask.tsx # Prompt templates +β”‚ β”‚ └── ... +β”‚ β”œβ”€β”€ config/ # Configuration utilities +β”‚ β”œβ”€β”€ icons/ # SVG icons (SVGR webpack loader) +β”‚ β”œβ”€β”€ locales/ # i18n translations (21 languages) +β”‚ β”‚ β”œβ”€β”€ en.ts +β”‚ β”‚ β”œβ”€β”€ cn.ts +β”‚ β”‚ └── ... +β”‚ β”œβ”€β”€ masks/ # Prompt template definitions +β”‚ β”‚ β”œβ”€β”€ build.ts # Mask builder script +β”‚ β”‚ β”œβ”€β”€ cn.json +β”‚ β”‚ β”œβ”€β”€ en.json +β”‚ β”‚ └── ... +β”‚ β”œβ”€β”€ mcp/ # Model Context Protocol +β”‚ β”œβ”€β”€ store/ # Zustand state stores +β”‚ β”œβ”€β”€ styles/ # Global SCSS +β”‚ β”‚ β”œβ”€β”€ globals.scss +β”‚ β”‚ β”œβ”€β”€ markdown.scss +β”‚ β”‚ └── ... +β”‚ └── utils/ # Utility functions +β”‚ β”œβ”€β”€ chat.ts +β”‚ β”œβ”€β”€ model.ts +β”‚ └── ... +β”œβ”€β”€ docs/ # Documentation (multilingual) +β”œβ”€β”€ public/ # Static assets +β”‚ β”œβ”€β”€ masks.json # Generated mask templates +β”‚ β”œβ”€β”€ serviceWorker.js # PWA service worker +β”‚ └── ... +β”œβ”€β”€ scripts/ # Build and deployment scripts +β”‚ β”œβ”€β”€ init-proxy.sh +β”‚ └── ... +β”œβ”€β”€ src-tauri/ # Tauri desktop app +β”‚ β”œβ”€β”€ src/ +β”‚ β”‚ └── main.rs # Rust entry point +β”‚ β”œβ”€β”€ tauri.conf.json # Tauri configuration +β”‚ └── Cargo.toml # Rust dependencies +β”œβ”€β”€ test/ # Jest tests +β”‚ β”œβ”€β”€ model-available.test.ts +β”‚ β”œβ”€β”€ model-provider.test.ts +β”‚ └── ... +β”œβ”€β”€ .github/ # GitHub workflows +β”‚ └── workflows/ +β”‚ β”œβ”€β”€ app.yml # Desktop builds +β”‚ β”œβ”€β”€ docker.yml # Docker publishing +β”‚ └── test.yml # CI tests +β”œβ”€β”€ Dockerfile # Production Docker image +β”œβ”€β”€ docker-compose.yml # Docker compose config +β”œβ”€β”€ next.config.mjs # Next.js configuration +β”œβ”€β”€ tsconfig.json # TypeScript configuration +β”œβ”€β”€ jest.config.ts # Jest configuration +β”œβ”€β”€ package.json # Node dependencies +└── .env.template # Environment variables template +``` + +--- + +## Development Workflows + +### Initial Setup + +```bash +# 1. Install dependencies (requires Node.js >= 18) +yarn install + +# 2. Create .env.local file +cp .env.template .env.local + +# 3. Add your API keys to .env.local +OPENAI_API_KEY=sk-xxxx +# Add other provider keys as needed + +# 4. Start development server +yarn dev +``` + +### Build Modes + +NextChat supports three distinct build modes: + +#### 1. Standalone (Default Web Server) +```bash +yarn build +yarn start +``` + +#### 2. Export (Static Export for Desktop) +```bash +yarn export +``` + +#### 3. Desktop App +```bash +yarn app:dev # Development with hot reload +yarn app:build # Production desktop build +``` + +### Development Scripts + +```json +{ + "dev": "concurrently \"yarn mask:watch\" \"next dev\"", + "build": "yarn mask && BUILD_MODE=standalone next build", + "export": "yarn mask && BUILD_MODE=export BUILD_APP=1 next build", + "app:dev": "concurrently \"yarn mask:watch\" \"yarn tauri dev\"", + "app:build": "yarn mask && yarn tauri build", + "mask": "npx tsx app/masks/build.ts", + "mask:watch": "npx watch \"yarn mask\" app/masks", + "test": "node --no-warnings --experimental-vm-modules jest --watch", + "test:ci": "node --no-warnings --experimental-vm-modules jest --ci", + "lint": "next lint" +} +``` + +### Git Workflow + +**Current Branch**: `claude/claude-md-mixl6cqn5jcf5sj4-019FCbQH47JyEPbiuunRgWXM` + +**Important**: All development should occur on feature branches starting with `claude/` and ending with the session ID. + +### Pre-commit Hooks + +Husky + lint-staged runs on every commit: +- Auto-fixes ESLint issues in `/app/**/*` +- Formats with Prettier +- Removes unused imports + +--- + +## Key Conventions + +### Code Style + +#### TypeScript +- **Strict mode**: Enabled +- **Path aliases**: Use `@/` for absolute imports (maps to project root) +- **Indentation**: 2 spaces +- **Quotes**: Double quotes +- **Trailing commas**: Required + +#### File Naming +- **Components**: PascalCase with `.tsx` extension + - Example: `ChatMessage.tsx` +- **Styles**: `component-name.module.scss` + - Example: `chat-message.module.scss` +- **API routes**: `route.ts` (Next.js App Router convention) +- **Types**: `typing.ts` or inline type definitions +- **Utilities**: camelCase with `.ts` extension + +#### Import Patterns +```typescript +// Use absolute imports via @ alias +import { ChatMessage } from "@/app/components/chat"; +import { useAccessStore } from "@/app/store/access"; + +// Dynamic imports for heavy components +const ModelConfigList = dynamic( + () => import("./model-config").then((mod) => mod.ModelConfigList), + { loading: () =>
Loading...
} +); +``` + +### Component Structure + +```typescript +// Standard component pattern +import styles from "./component.module.scss"; +import { useState } from "react"; + +export function MyComponent(props: MyComponentProps) { + const [state, setState] = useState(); + + return ( +
+ {/* Component JSX */} +
+ ); +} +``` + +### State Management + +```typescript +// Zustand store pattern +import { create } from "zustand"; +import { persist } from "zustand/middleware"; + +interface MyState { + value: string; + setValue: (value: string) => void; +} + +export const useMyStore = create()( + persist( + (set, get) => ({ + value: "", + setValue: (value) => set({ value }), + }), + { + name: "my-store", + } + ) +); +``` + +### API Route Pattern + +```typescript +// app/api/[provider]/[...path]/route.ts +import { NextRequest, NextResponse } from "next/server"; + +export const runtime = "edge"; // Use Edge Runtime + +export async function POST(req: NextRequest) { + // Handle request + return NextResponse.json({ data }); +} +``` + +### Styling Conventions + +- Use **CSS Modules** for component styles +- Global styles in `/app/styles/globals.scss` +- Follow BEM-like naming within modules +- Use SCSS variables for theming + +```scss +// component.module.scss +.container { + display: flex; + + &__header { + padding: 1rem; + } + + &__content { + flex: 1; + } +} +``` + +--- + +## Common Tasks & Patterns + +### Adding a New AI Provider + +1. **Create provider client** in `/app/client/platforms/` +```typescript +// app/client/platforms/newprovider.ts +export class NewProviderApi implements LLMApi { + // Implement LLMApi interface +} +``` + +2. **Add API handler** in `/app/api/` +```typescript +// app/api/newprovider.ts +export function makeNewProviderHandler() { + // Handler implementation +} +``` + +3. **Update model configuration** in `/app/store/config.ts` + +4. **Add environment variables** to `.env.template` + +5. **Update locales** in `/app/locales/` for all languages + +### Adding a Translation Key + +1. **Add to English locale** (primary): `/app/locales/en.ts` +```typescript +const en = { + MyFeature: { + Title: "My Feature", + Description: "Feature description", + }, +}; +``` + +2. **Add to other locales**: `/app/locales/cn.ts`, etc. + +3. **Use in component**: +```typescript +import Locale from "@/app/locales"; + +function MyComponent() { + return

{Locale.MyFeature.Title}

; +} +``` + +### Creating a Prompt Template (Mask) + +1. **Add template** to `/app/masks/cn.json` or `/app/masks/en.json` +```json +{ + "avatar": "icon-emoji", + "name": "My Template", + "context": [ + { + "role": "system", + "content": "You are a helpful assistant." + } + ], + "modelConfig": { + "model": "gpt-4", + "temperature": 0.7 + } +} +``` + +2. **Build masks**: `yarn mask` + +3. **Output**: Generated to `/public/masks.json` + +### Adding a New Component + +1. **Create component file**: `/app/components/my-feature.tsx` +```typescript +import styles from "./my-feature.module.scss"; + +export function MyFeature() { + return
Content
; +} +``` + +2. **Create styles**: `/app/components/my-feature.module.scss` + +3. **Add to route** (if needed) in `/app/components/home.tsx` + +### Working with Environment Variables + +1. **Add to `.env.template`** with documentation +2. **Access in API routes**: +```typescript +const apiKey = process.env.MY_PROVIDER_API_KEY; +``` +3. **Access in client** (must be prefixed with `NEXT_PUBLIC_`): +```typescript +const publicValue = process.env.NEXT_PUBLIC_MY_VALUE; +``` + +### Model Context Protocol (MCP) Integration + +**Enable MCP**: Set `ENABLE_MCP=true` in environment variables + +**MCP files location**: `/app/mcp/` + +**Server actions**: Use Next.js server actions for client management + +--- + +## Testing Guidelines + +### Test Configuration + +- **Framework**: Jest 29.7.0 +- **Environment**: jsdom (for React components) +- **Config**: `/jest.config.ts` +- **Setup**: `/jest.setup.ts` + +### Running Tests + +```bash +# Watch mode (development) +yarn test + +# CI mode (one-time run) +yarn test:ci +``` + +### Test File Locations + +All tests in `/test/` directory: +- `model-available.test.ts` - Model availability checks +- `model-provider.test.ts` - Provider validation +- `vision-model-checker.test.ts` - Vision model capabilities + +### Writing Tests + +```typescript +import { render, screen } from "@testing-library/react"; +import { MyComponent } from "@/app/components/my-component"; + +describe("MyComponent", () => { + it("should render correctly", () => { + render(); + expect(screen.getByText("Expected Text")).toBeInTheDocument(); + }); +}); +``` + +### Code Quality + +**ESLint**: Next.js core web vitals + Prettier + unused-imports +```bash +yarn lint +``` + +**Prettier**: Configured for 80-char width, 2-space tabs +```bash +npx prettier --write "app/**/*.{ts,tsx}" +``` + +--- + +## Deployment + +### Docker (Recommended) + +```bash +# Pull image +docker pull yidadaa/chatgpt-next-web + +# Run with environment variables +docker run -d -p 3000:3000 \ + -e OPENAI_API_KEY=sk-xxxx \ + -e CODE=your-password \ + yidadaa/chatgpt-next-web +``` + +**With proxy**: +```bash +docker run -d -p 3000:3000 \ + -e OPENAI_API_KEY=sk-xxxx \ + -e PROXY_URL=http://localhost:7890 \ + yidadaa/chatgpt-next-web +``` + +**With MCP enabled**: +```bash +docker run -d -p 3000:3000 \ + -e OPENAI_API_KEY=sk-xxxx \ + -e ENABLE_MCP=true \ + yidadaa/chatgpt-next-web +``` + +### Docker Compose + +Two profiles available: +1. **no-proxy**: Direct connection +2. **proxy**: With PROXY_URL support + +```bash +docker-compose up -d +``` + +### Vercel + +1. Click "Deploy with Vercel" button +2. Set environment variables: + - `OPENAI_API_KEY` (required) + - `CODE` (access password, optional) +3. Deploy + +**Important**: Fork the repository (don't just clone) to enable automatic updates. + +### Cloudflare Pages + +See documentation: `/docs/cloudflare-pages-en.md` + +### Self-Hosted + +```bash +# Install dependencies +yarn install + +# Build +yarn build + +# Start +yarn start +``` + +### Desktop App Distribution + +```bash +# Build for current platform +yarn app:build + +# Output location: src-tauri/target/release/ +``` + +--- + +## AI Assistant Guidelines + +### When Working with This Codebase + +#### 1. **Always Read Before Modifying** +- Never propose changes to code you haven't read +- Use the Read tool to examine files before suggesting modifications +- Understand the existing patterns and conventions + +#### 2. **Follow Existing Patterns** +- This codebase has established patterns for: + - Multi-provider API integration + - State management with Zustand + - Component structure and styling + - Internationalization +- Match the existing code style and architecture + +#### 3. **Respect the Build System** +- **Masks must be built**: Run `yarn mask` after modifying `/app/masks/` +- **Three build modes**: Be aware of standalone/export/desktop differences +- **Environment variables**: Use `.env.local` for development + +#### 4. **Internationalization is Critical** +- Always update ALL language files in `/app/locales/` when adding UI text +- Primary language is English (`en.ts`) +- 21 languages must be kept in sync + +#### 5. **Multi-Provider Architecture** +- When adding features, consider impact on all 15+ providers +- Test with multiple providers when possible +- Use the unified `getClientApi()` interface + +#### 6. **Security Considerations** +- **API keys**: Never expose in client-side code +- **Environment variables**: Server-side only unless prefixed `NEXT_PUBLIC_` +- **CORS**: Be mindful of cross-origin restrictions +- **Input validation**: Always validate user input + +#### 7. **Performance Best Practices** +- Use dynamic imports for heavy components +- Leverage Next.js Edge Runtime for API routes +- Minimize bundle size (target: ~100kb first screen) +- Use streaming for real-time responses + +#### 8. **Testing Requirements** +- Add tests for new utilities and core logic +- Place tests in `/test/` directory +- Run `yarn test:ci` before committing + +#### 9. **Documentation** +- Update README.md for user-facing changes +- Update this CLAUDE.md for architectural changes +- Add JSDoc comments for complex functions + +#### 10. **Common Pitfalls to Avoid** +- ❌ Don't bypass the mask build process +- ❌ Don't hardcode API endpoints (use environment variables) +- ❌ Don't add dependencies without checking bundle size impact +- ❌ Don't modify provider-specific code without testing all providers +- ❌ Don't skip internationalization +- ❌ Don't use `any` types in TypeScript +- ❌ Don't add inline styles (use CSS Modules) + +### Search Strategies + +#### Finding Provider-Related Code +```bash +# Search for provider implementations +app/client/platforms/*.ts + +# Search for API handlers +app/api/*.ts + +# Search for provider configuration +app/store/config.ts +``` + +#### Finding UI Components +```bash +# Search in components directory +app/components/*.tsx + +# Search for specific feature +grep -r "feature-name" app/components/ +``` + +#### Finding Localization Strings +```bash +# Search in locales +app/locales/*.ts +``` + +### Code Modification Checklist + +When making changes, verify: +- [ ] Read all relevant files first +- [ ] Followed existing code patterns +- [ ] Updated all 21 language files (if UI text changed) +- [ ] Ran `yarn mask` (if masks modified) +- [ ] Tested in development mode (`yarn dev`) +- [ ] Ran linter (`yarn lint`) +- [ ] Ran tests (`yarn test:ci`) +- [ ] Updated documentation (if needed) +- [ ] Checked bundle size impact (if dependencies added) +- [ ] Tested with multiple AI providers (if API-related) + +### Understanding the Request Flow + +**Web App Request Flow**: +``` +User Browser + ↓ +Next.js Frontend (React) + ↓ +app/components/chat.tsx + ↓ +app/client/api.ts (getClientApi) + ↓ +app/api/[provider]/route.ts (Edge Runtime) + ↓ +AI Provider API (OpenAI, Claude, etc.) +``` + +**Desktop App Request Flow**: +``` +Tauri Desktop App + ↓ +Next.js Frontend (Static Export) + ↓ +Tauri Fetch API (Rust) + ↓ +AI Provider API (Direct) +``` + +--- + +## Important Files Reference + +### Configuration Files + +| File | Purpose | +|------|---------| +| `/next.config.mjs` | Next.js configuration, webpack, build modes, rewrites | +| `/tsconfig.json` | TypeScript config, path aliases (`@/*`) | +| `/jest.config.ts` | Jest test configuration | +| `/.env.template` | Environment variables template (all available vars) | +| `/.env.local` | Local environment (gitignored, create manually) | +| `/package.json` | Dependencies and scripts | +| `/src-tauri/tauri.conf.json` | Tauri desktop app configuration | + +### Entry Points + +| File | Purpose | +|------|---------| +| `/app/layout.tsx` | Root HTML structure, metadata, analytics | +| `/app/page.tsx` | Root page, renders `` | +| `/app/components/home.tsx` | Main app shell with client-side routing | +| `/app/api/[provider]/[...path]/route.ts` | Dynamic API route handler | +| `/src-tauri/src/main.rs` | Tauri/Rust desktop entry point | + +### Core Logic + +| File | Purpose | +|------|---------| +| `/app/client/api.ts` | Unified API client interface | +| `/app/client/platforms/*.ts` | Individual provider implementations | +| `/app/store/*.ts` | Zustand state stores | +| `/app/utils/chat.ts` | Chat utility functions | +| `/app/utils/model.ts` | Model configuration utilities | + +### UI Components + +| File | Purpose | +|------|---------| +| `/app/components/chat.tsx` | Main chat interface | +| `/app/components/settings.tsx` | Settings panel | +| `/app/components/mask.tsx` | Prompt template management | +| `/app/components/model-config.tsx` | Model configuration UI | +| `/app/components/sidebar.tsx` | Application sidebar | + +### Styling + +| File | Purpose | +|------|---------| +| `/app/styles/globals.scss` | Global styles | +| `/app/styles/markdown.scss` | Markdown rendering styles | +| `/app/styles/animation.scss` | Animation utilities | + +### Internationalization + +| File | Purpose | +|------|---------| +| `/app/locales/en.ts` | English (primary) | +| `/app/locales/cn.ts` | Simplified Chinese | +| `/app/locales/tw.ts` | Traditional Chinese | +| `/app/locales/*.ts` | 18 other languages | + +### Build & Deployment + +| File | Purpose | +|------|---------| +| `/Dockerfile` | Production Docker image (multi-stage) | +| `/docker-compose.yml` | Docker compose configuration | +| `/.github/workflows/*.yml` | CI/CD workflows | +| `/scripts/*.sh` | Build and deployment scripts | + +--- + +## Environment Variables Quick Reference + +### Required +- `OPENAI_API_KEY`: OpenAI API key + +### Optional Provider Keys +- `ANTHROPIC_API_KEY`: Claude API key +- `GOOGLE_API_KEY`: Gemini API key +- `DEEPSEEK_API_KEY`: DeepSeek API key +- `BAIDU_API_KEY` & `BAIDU_SECRET_KEY`: Baidu API +- `BYTEDANCE_API_KEY`: ByteDance API +- `ALIBABA_API_KEY`: Alibaba Cloud API +- `CHATGLM_API_KEY`: ChatGLM API +- `SILICONFLOW_API_KEY`: SiliconFlow API +- `AI302_API_KEY`: 302.AI API + +### Feature Flags +- `CODE`: Access password (comma-separated) +- `DISABLE_GPT4`: Set to `1` to disable GPT-4 +- `HIDE_USER_API_KEY`: Set to `1` to hide user API key input +- `ENABLE_BALANCE_QUERY`: Set to `1` to enable balance queries +- `DISABLE_FAST_LINK`: Set to `1` to disable URL settings parsing +- `ENABLE_MCP`: Set to `true` to enable Model Context Protocol + +### Customization +- `CUSTOM_MODELS`: Add/remove models (e.g., `+llama,-gpt-3.5-turbo`) +- `DEFAULT_MODEL`: Change default model +- `VISION_MODELS`: Add vision capabilities to models +- `DEFAULT_INPUT_TEMPLATE`: Customize input preprocessing template + +### URLs & Endpoints +- `BASE_URL`: Override OpenAI API base URL +- `PROXY_URL`: HTTP proxy URL +- `ANTHROPIC_URL`: Custom Claude API URL +- `GOOGLE_URL`: Custom Gemini API URL +- `WHITE_WEBDAV_ENDPOINTS`: Allowed WebDAV endpoints + +--- + +## Recent Changes & Updates + +### v2.16.1 (Current) +- Support for GPT-5 models +- Use `max_completion_tokens` parameter +- MCP (Model Context Protocol) support +- Real-time chat with WebRTC audio + +### Recent Commits +- `45f7b91`: /all +- `995bef7`: Merge PR #6599 - Add support for GPT5 +- `38ac502`: Add support for GPT5 +- `0511808`: Use max_completion_tokens +- `42eff64`: Use max_completion_tokens + +--- + +## Additional Resources + +### Documentation +- [Deployment with Cloudflare](./docs/cloudflare-pages-en.md) +- [FAQ (English)](./docs/faq-en.md) +- [How to add translation](./docs/translation.md) +- [User Manual (Chinese)](./docs/user-manual-cn.md) + +### External Links +- [NextChat Official Site](https://nextchat.club) +- [Web App Demo](https://app.nextchat.club) +- [GitHub Repository](https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web) +- [Desktop App Releases](https://github.com/Yidadaa/ChatGPT-Next-Web/releases) + +--- + +## Contact & Support + +- **Enterprise Inquiries**: business@nextchat.dev +- **Issues**: GitHub Issues +- **Discussions**: GitHub Discussions + +--- + +**End of CLAUDE.md** + +*This document should be updated whenever significant architectural changes are made to the codebase.*