As a heavy user of Cursor, I rely on it daily for both coding requests and research tasks.
While it has become a critical part of my workflow, I find the inability to control AI response
formatting increasingly frustrating. This limitation disrupts productivity, especially when
alternating between concise coding queries and more exploratory, creative tasks.
One of my biggest annoyances is being spammed with long bulleted or numbered lists when all I want
is a short, well-written paragraph in prose. To address this, I propose a toolbar with customizable
options for response length and formatting style.
Key Features
Response Length Options:
- One-liner: Quick, to-the-point solutions.
- Concise: Minimal detail, focused on key points.
- Detailed: Full explanations with context.
- Comprehensive: In-depth responses with alternatives.
- Custom Token Limit: Fine-tune responses to specific needs.
Formatting Style Toggles:
Easily switch between these styles based on the task:
- Bullet Points
- Numbered Lists
- Flowchart Style
- Q&A Format
- Table Format
- Paragraph Style
- Code-Focused (minimizes prose)
Additional Style Request
I also propose including style toggles like “bash format,” which I use as shorthand to request
that the LLM send me responses in plain text. This avoids formatting issues that occur when
bulleted or numbered lists are sent, making it easier to copy-paste responses seamlessly
into my workflow.
Implementation
A small, unobtrusive toolbar near the AI interaction panel could allow users to toggle these
preferences, sending them along with the query to dynamically tailor responses.
Why This Matters
This feature would streamline workflows, minimize irrelevant formatting, and make Cursor
adaptable to varied use cases—whether debugging code, conducting research, or brainstorming
creative ideas.
Would Cursor’s team consider implementing this? It would greatly enhance the tool’s versatility
and usability for dedicated users like me.