ChatALL

Sends a single user prompt to 30+ AI bots simultaneously through a debounced message queue system that batches updates and persists state to IndexedDB. Uses Vuex mutations to coordinate state changes across multiple bot instances, with IPC handlers managing bot-specific connection protocols (API keys, web sessions, proxy settings). The queue.js module implements debounced persistence to prevent excessive database writes while maintaining consistency across the Electron main and renderer processes.

Renders bot responses in configurable 1, 2, or 3-column layouts using Vue.js components with CSS Grid, enabling visual comparison of identical prompts across different models. The UI layer (App.vue, SettingsModal.vue) manages column count state through Vuex mutations, with responsive design adapting to window resize events. Each column independently streams responses from its assigned bot, with scroll synchronization and message threading support via the message display system.

Organizes messages into threaded conversations with support for branching (multiple responses to the same prompt). Each message is linked to a parent message via a thread ID, enabling tree-like conversation structures. The message display system renders threads with visual indentation and parent-child relationships. Users can view the full conversation history or focus on a specific thread. Threading is persisted in IndexedDB with the messages and threads tables.

Provides dark and light UI themes with automatic detection of system theme preference via native OS APIs. The main process (background.js) queries the system theme using Electron's nativeTheme API and communicates it to the renderer via IPC. Users can override the system preference with manual theme selection, which is persisted in Vuex state. Theme switching is instant and affects all UI components via CSS variables.

Provides keyboard shortcuts for common actions (send message, new chat, switch bots, etc.) with customizable hotkey bindings. Shortcuts are defined in configuration and registered with the Electron main process, enabling global hotkeys that work even when the window is not focused. The UI displays shortcut hints next to buttons. Hotkey bindings are persisted in Vuex state and can be customized via settings.

Checks for new application versions on startup and periodically in the background, with user-facing notifications for available updates. The update system compares the current version (from package.json) with the latest release on GitHub, displaying a notification if an update is available. Users can manually trigger update checks via settings. Update installation requires manual download and installation; no automatic patching.

Integrates LangChain library to support AI models without native SDKs, using LangChain's unified interface for prompt execution and response parsing. LangChain abstracts provider-specific APIs (OpenAI, Anthropic, Hugging Face, etc.) into a common interface, enabling ChatALL to support models beyond those with dedicated integrations. Bot implementations can use LangChain's LLM classes, chains, and agents for complex prompt workflows. LangChain integration adds ~200-500ms overhead per request due to abstraction layers.

Supports OpenAI-compatible APIs (e.g., local LLMs running on OpenAI-compatible servers, Azure OpenAI) by allowing users to configure custom API endpoints. The OpenAI bot implementation accepts a custom base URL parameter, enabling connection to any OpenAI-compatible server. This enables users to run local models (via llama.cpp, vLLM, etc.) or use alternative providers (Azure, Replicate) without modifying code. API key and endpoint are persisted in bot configuration.

Defines a Bot base class hierarchy that abstracts heterogeneous AI service APIs (OpenAI, Anthropic, Bing, LangChain, Chinese services) into a unified interface with send(), onMessage(), and onError() methods. Each bot subclass implements provider-specific authentication (API keys, OAuth, web scraping), request formatting, and response parsing. The system uses LangChain integration for models without native SDKs, with fallback to direct HTTP calls for services like Bing Chat. Bot configuration is persisted in Vuex state and IndexedDB, enabling dynamic bot registration without code changes.

Persists all chat sessions, messages, and bot responses to IndexedDB using the Dexie ORM library, with tables for chats, messages, and threads. The Vuex store coordinates state changes through mutations that trigger debounced writes to IndexedDB via queue.js, preventing excessive database operations. Chat history is loaded on application startup and kept in-memory for fast access, with lazy-loading of older messages when users scroll. All data is stored locally on the user's machine; no cloud synchronization.

Implements internationalization (i18n) using Vue.js i18n plugin with JSON translation files for 10 languages (English, Chinese, Japanese, Korean, Spanish, French, German, Italian, Russian, Vietnamese). The i18n/index.js module loads locale files dynamically based on system language detection or user preference, with fallback to English. All UI strings, button labels, and error messages are externalized to translation files, enabling language switching without application restart.

Packages ChatALL as a native desktop application using Electron, combining a Node.js main process with a Chromium renderer running Vue.js. The main process (background.js) handles window management, IPC communication, system theme detection, proxy configuration, and cookie access via native APIs. The renderer process (App.vue) runs the Vue.js UI and communicates with the main process via IPC for privileged operations. Supports Windows, macOS, and Linux with platform-specific installers.

Allows users to save frequently-used prompts to a local prompt library, with tagging and search capabilities. Saved prompts are persisted in Vuex state and IndexedDB, enabling quick insertion into the chat input. The SettingsModal.vue component provides UI for managing prompt collections. Prompts can be organized by category tags and searched by keyword, reducing repetitive typing for common queries.

Manages authentication credentials for 30+ AI services (API keys, OAuth tokens, usernames/passwords) with secure storage in IndexedDB and in-memory caching. The bot system handles provider-specific auth flows: API key validation for OpenAI/Anthropic, OAuth for some services, and web session management for browser-based bots. Credentials are validated on bot initialization and cached in memory to avoid repeated authentication. Failed authentication triggers error handling with user-facing prompts to re-enter credentials.

Allows users to configure HTTP/HTTPS proxy settings for all bot API requests, with support for proxy authentication (username/password). Proxy configuration is persisted in IndexedDB and loaded on application startup. The main process (background.js) provides IPC handlers to read and save proxy settings, which are then applied to all HTTP clients (axios, node-fetch) used by bot implementations. Supports both system proxy detection and manual proxy configuration.

Renders bot responses in real-time as they stream from the API, using Vue.js reactive data binding to update the UI incrementally. Each bot's response is streamed to a message object in Vuex state, with the UI component re-rendering on each chunk received. The message display system handles markdown rendering, code syntax highlighting, and text formatting. Streaming is non-blocking; the UI remains responsive while responses are being received.