Another option would be adding an Information icon that lets radar detect completely unopened Explorables.
Right now, Solve Explorable behaves very differently depending on those two cases. If the input is a completely unopened Explorable, the instruction first moves toward it in a synchronized way, and there’s no way to set a timeout or run it asynchronously. Because of that, the bot can sometimes get attacked by bugs.
At the moment, there’s no way to check an Explorable’s state in advance, so it’s impossible to predict how the Solve Explorable instruction will behave.
