Most agentic SWE research is obsessed with the syntax. We spend months training models to write perfect Python functions or elegant Rust crates, yet we still rely on humans to babysit the compiler. We treat the environment as a given, a static stage where the code performs.