Stopping Play Mode for a one-line fix throws away the match and the bug you finally reproduced. Satura Space patches edited method bodies into the running session instead.
The usual loop is: stop Play Mode, compile, reload the domain, press Play, replay your way back. That's a minute or two on a mid-size project, and an agent that writes five fixes pays it five times.
With hot-reload armed, the edited script compiles and its method bodies are patched into the running Mono domain. Unity stays in Play Mode, nothing is torn down, and the new code runs the next time the method is called.
Tune a jump curve and watch it change in a game that never stopped.
Assembly-CSharp, your asmdefs and embedded packages.Awake or Start only affects objects spawned afterwards.With unity_exec on the Unity MCP server an agent runs C# in the live session to spawn objects, change values and read state, through Roslyn in memory with no recompile or domain reload. The demo above drops 180 cubes into a game that never paused.
Unity 6Mono scripting backend. Tested day to day on 6000.6 and 6000.3.
Method bodiesExisting methods in your own assemblies. Signature and type changes recompile.
Fully preservedNo domain reload, no Play Mode exit. Scene objects and connections survive.
hot_reload=trueOn unity_open_scene: an agent arms the same toggle the Player panel exposes.
Method bodies, during Play Mode, on Unity 6 (Mono), in Assembly-CSharp, asmdefs and embedded packages. New fields, methods and types need a normal recompile.
No, it needs Mono, which the editor uses. Your IL2CPP player builds are unaffected.
Patch the running game, keep the state, and let your agents do the same.