Terminal vs IDE: Where do you prefer your AI coding agents?

Hey everyone! I’m launching Ersinis tomorrow (an autonomous multi-agent developer system), and I had a huge debate with myself during the build process.

Tools like Cursor and Copilot live directly in the IDE or browser. But I decided to build Ersinis as a terminal-native agent because I feel the terminal is where the real "magic" and file executions happen.

Do you prefer your AI assistants strictly inside your code editor (like VS Code/Cursor), or do you like having a terminal-based autonomous agent that acts like a standalone team member?

Would love to hear your thoughts before my launch tomorrow!