Torque is hard to understand. If anyone tells you otherwise, they're a liar and they want something from you.
Even physicists who totally get it now had a hard time learning it. It's tricky. It does unexpected things sometimes.

Fundamentally, a torque is the rotational equivalent of a force. A force causes an acceleration in a straight line, and a torque causes a spinning acceleration.

In physics, we describe a lot of things with vectors. Vectors have a magnitude and a direction. An arrow has a length and a direction. We readily use arrows as symbols for vectors, with the magnitude of the vector corresponding to the length of the arrow, and direction is same for both.

With force, if your butt pushes on the chair with your weight, we can describe that force with a vector located at your butt and pointing down. It's length would correspond to your weight. It's pretty straight forward. Your butt pushes down, so the arrow points down.

With torque it's trickier. A thing starts spinning. Part of it is moving in every direction (except the directions parallel to the axis of rotation). There's no unique direction which indicates what the torque is doing. The axis of rotation would be the ideal place to put the vector, but that exact point doesn't move at all.
So we make an executive decision.
We decide to point the torque vector along the axis of rotation. This isn't fully solving the problem, though. The axis goes both ways. I mean... there's a spinning thing, and the axis goes through it. So depending on which side we attach the torque arrow, it could be flipped 180 in direction.
Another executive decision.
We still need to link the direction of spinning to the arrow, too. It could be going clockwise or counterclockwise (from a defined coordinate perspective). We can link the direction of rotation with the 180 flip of the arrow and clean up both problems if we simply pick a standardized way to link the direction of the spinning to the direction of the arrow which describes the vector which describes the spinning.

We use the right hand rule.
I can get into this in more detail if you like. For now, it's enough to say
if I open my right palm with my thumb out,
then close my fingers, keeping my thumb out,
my thumb points in the direction of the vector which describes the spinning of my fingers.

(Notice that if you do this with your left hand, it's a mirror image, and flips the direction of the thumb relative to the rotation. We could have used this rule in our 2nd executive decision. Which we use isn't relevant, so long as all of our math and physics is consistent. The right hand rule comes up a lot, so just go with it... even if you're left-handed.)

Here's a somewhat more involved explanation of torque as pertains to not cars.