Agentic coding ruined normal IDEs like VS Code, Zed, and PyCharm for me, and I don’t think I can go back

A developer shares how agentic coding tools have permanently changed their expectations of IDEs. After years of incremental improvements like faster syntax highlighting and AI autocomplete, experiencing a true agentic workflow — where the AI autonomously plans tasks, navigates multiple modules, handles dependency errors, and reruns tests without human intervention — made traditional IDEs like VS Code, Zed, and PyCharm feel fundamentally broken. The key difference is that legacy IDEs treat AI as a guest extension requiring manual copy-paste loops, while agent-first tools like Google Antigravity integrate terminal, browser, and file management into a single execution loop. The author argues we are entering a post-text-editor era where the next-gen IDE wins by acting as a native peer, not by offering marginally better syntax highlighting.