My workflow was to use Claude/ChatGPT for security research assistance, basically a Google step-in, and Cursor in manual mode to generate and modify code. Wasn't really too fond of agent mode in Cursor but that now seems like the standard, that or Claude Code generating heaps of code?
So what does your workflow look like in your day-to-day? Anything other than Cursor/CC + some LLM provider?