Ollama is running. The model loaded. You’ve confirmed it generates clean code from the command line. Now you open a coding session and immediately hit the workflow problem: you type a prompt in one terminal, copy the output, switch to your editor, paste it, check the line numbers, realize …