Hey everyone! I’m a big fan of the Cursor chat—it’s been a game-changer for my workflow. I’ve got an idea for a feature that could make typing prompts even smoother, and I’d love to hear your thoughts! When I’m typing prompts in the chat sidebar—whether to ask a question, edit code, or chat with an agent like Claude-3.7-Sonnet—I often repeat patterns or phrases from earlier in the conversation. It’d be awesome to have a way to speed this up without typing everything out manually.
As I type in the prompt input field (the rectangular box with the ‘Ask’ button and ‘Send’ arrow), I’d like to see a faint, semi-transparent ‘shadow’ of potential autocomplete tokens appear ahead of my cursor. For example, if I type ‘Can you help me with,’ a shadow might show ‘debugging this code’ in light gray. I’d love to press the Tab key to cycle through different suggestions (e.g., next word or phrase) and the right arrow key to autofill the current suggestion into the input box. If I don’t want the suggestion, I can just keep typing to ignore it.
The suggestions should pull context from the ongoing chat. For instance, if I’ve been discussing Python loops, typing ‘How do I’ might suggest ‘optimize this loop’ instead of something generic.
It’d cut down on typing time, especially for longer prompts or repetitive tasks. Context-aware suggestions would make my prompts sharper and more aligned with the conversation. It’d feel intuitive, like autocomplete in code editors, but tailored for chatting with the AI.
I think this could make the chat sidebar even more powerful and fun to use. Thanks for checking out my idea—I’m excited to see what the Cursor team and community think!