Better AI Chat: Message Queue

**Better AI Chat: Message Queue** adds a message queue to your AI chat workflow so you can line up multiple prompts and send them one-by-one, automatically, right after each response finishes. It runs entirely in your browser, modifies only the chat interface, and never sends your data to any external server. ### What it does (today) Right now, Better AI Chat focuses on improving the ChatGPT web experience at `chatgpt.com` / `chat.openai.com` by injecting a small, unobtrusive UI above the composer and managing a local message queue. * **Queue messages while a reply is generating** When a response is generating, press **Alt+Enter** to add your current draft to the queue instead of sending it immediately. * **Automatic sending after each response** Once the current AI response finishes, the extension automatically moves to the next queued message and sends it for you – no manual copy/paste, no babysitting the UI. * **Inline queue preview bar** A slim queue bar sits above the input box, showing how many messages are queued, with quick access to expand, re-order, edit, or delete those messages. * **Full queue manager UI** Expand the queue to see each message with: • Numbered order (what sends next) • Two-line preview of the text • Buttons to move up/down, edit, or delete individual messages * **Popup editor** Open a popup editor from the queue bar (“Edit in Popup”) to draft longer, multi-paragraph prompts in a distraction-free textarea, then insert them back into the chat input or send/queue directly. * **Per-chat queues, saved locally** Queues are stored per conversation using `localStorage`, so each chat can have its own queued messages, and everything stays on your machine. * **Typing-aware sending** The extension detects when you’re actively typing and waits until you pause before sending the next queued message, to avoid interrupting your edits. ### Why it’s useful * **Batch your thinking** – Write out a whole sequence of prompts and let them run automatically instead of constantly waiting and clicking “Send.” * **Stay in flow** – Perfect for long research sessions, code reviews, or brainstorming where you already know the next 3–10 questions you want to ask. * **Fewer mistakes** – Reordering and editing queued prompts before they send helps you avoid “oops, wrong prompt in the wrong chat.” ### How it works under the hood * Injects a content script on `https://chatgpt.com/*` and `https://chat.openai.com/*` that: * Detects the chat input (textarea or contenteditable) and the Send button using resilient selectors that adapt to UI changes. * Watches the page for when the AI is generating a reply vs. idle using DOM attributes and visual indicators. * Intercepts **Alt+Enter** to push your current message text into a local queue and clear the input. * After generation finishes, loads the next queued message into the editor and triggers the Send button like a normal click. There is no background server, no remote API calls, and no data collection — everything is JavaScript running inside your browser tab. ### Compatibility & Roadmap > **Important:** At the moment, this extension **only** runs on the ChatGPT web interface (`chatgpt.com` and `chat.openai.com`). Better AI Chat: Message Queue is designed as a "make AI chat better" layer. The long-term vision is to bring similar queueing and workflow tools to other popular AI chat platforms in the future. This extension is an independent project and is **not affiliated with, endorsed by, or sponsored by OpenAI** or any other company. ### Who this extension is for * Power users who live in ChatGPT for coding, writing, or research * Students and researchers running structured experiment-style prompts * Content creators drafting long series of variations or A/B tests * Anyone who feels like they’re constantly stuck in “Send → wait → Send → wait” loops ### Support, feedback, and feature requests If you need help, run into bugs, or the extension stops working due to UI changes, you can contact me directly: * **LinkedIn:** search for **“Elliot Drel”** * **Discord:** **@flyingymnast** If you send me a feature request and it’s feasible for this extension, I’ll either add it or I’ll pay you **$20** as a thank-you for the idea.