Feature request for product/service
Cursor CLI
Describe the request
Feature Request: Cursor CLI Enhancements
Product/Service: Cursor CLI
Description of the request:
Hi team, I would like to request several improvements to Cursor CLI to bring it closer to the functionality offered by Claude Code and Cursor IDE:
-
Custom Slash Commands Folder (
/)- Allow users to create and manage a folder of custom slash commands.
- This would enable flexible workflows and personalized automation, similar to Claude Code’s customization features.
-
Memory Management
- Add persistent memory features to retain context across sessions.
- Provide commands to enable/disable memory or selectively clear it.
-
Model Selection Options
- Expand the available models in Cursor CLI (currently very limited).
- At minimum, offer half the model set of Cursor IDE, so users can experiment with different strengths.
- Example: While using Claude 4 in Cursor CLI, I noticed it is not the “thinking” model variant, which limits advanced reasoning.
-
Terminal Integration Improvements
- Fix issues where long-running commands cause cursor-agent to hang indefinitely.
- Improve robustness of terminal interactions, especially for background jobs and streaming logs.
-
Bug Fixes and Stability
- Address occasional freezes when CLI tries to process outputs that exceed certain lengths.
- Ensure smoother cancellation (e.g.,
Ctrl+Cshould reliably stop both the command and the agent).
-
Additional Features Inspired by CodeX / Claude Code:
- Plugin/extension hooks: allow community-built add-ons for advanced workflows.
- Improved output formatting: support for structured JSON/YAML output when requested, making automation easier.
Operating Systems:
- Windows 10/11
- macOS
- Linux