People have successfully dropped autonomous robots onto Mars before. Wouldn't be out of the question to do so again, except that instead of rovers, the robots could be programmed to open up crates (dropped at the same time, or by a previous or future mission) and assemble what's inside into a landing or launch pad.
Would take a lot of prep missions, and might not be worth the cost, but in theory it's doable. Curiosity massed 900 kg and Perseverance massed 1025 kg according to Wikipedia, so dropping a metric ton of supplies for the robot to assemble would, again in theory, be feasible. (In practice the equipment for landing the supply crates safely would take up a good chunk of the mass budget, but there should be something left over in the mass budget for the crates to contain).
If you have robots like this, why send humans in the first place then? You've already got the bots that can do pretty much everything a human could want to do and you don't have to get them back off Mars.
Robots specialized in assembling one machine are not general-purpose robots. I doubt they would be able to "do pretty much everything a human could want to do".
Now, if someone does perfect a general-purpose, AI-driven robot, then yes, sending that robot on the Mars mission instead of a human would make a lot of sense. That's kind of, kind of, what we've done already with the rovers. But until such a general-purpose robot is actually created, human beings are still the most versatile, adaptable entities that could go to Mars.
I doubt it would be practical to pump fuel to another rocket. The obvious thing to do is store the fuel in orbit with its own rocket engine. Then you only have to dock with it and turn on the engine.
