Is this 'sky monster' astronomically possible?

Question

The sky monster is a singular light in the skies of Droelep, a rather Earth-like planet in a system much like our solar system

For part of the year, it moves from horizon to horizon on a smooth line, no different to any star. But at a point it begins to wander like a planet, following a unique path in the sky. Later in the year it takes on a meteoric nature, appearing and disappearing from view wherever it pleases. But this phase soon ends and it returns to its planetary path, before it returns to its place with the stars to repeat the cycle over again

Is there any object out in space that could display such bizarre behaviour in the night sky?

This universe has no magic in its astronomy, and space is essentially the same as in reality

Answer 1

As stated in other answers, there isn't really a good natural candidate for this object.

But if you are willing to permit an artificial object, then there is something that comes close:

A large interstellar probe

Launched many thousands of years ago from a distant star-system, the interstellar probe is initially invisible, coasting through space at a measurable fraction of the speed of light.

Far out from your solar system it ignites its engine and begins to decelerate. In order to journey between star systems it would need an absurdly powerful engine, perhaps powered by nuclear fusion, antimatter, or some exotic physical processes that are currently unknown to us. Even so it would take a long time to decelerate. While it does this over months or years, it's engine is visible as a steadily brightening star in the sky. It would not appear to move much, as it is moving towards the solar system it is being observed from, meaning there would be relatively little angular deflection.

As it decelerates into the solar system, passing inside the orbit of comets and outer planets, it will now be moving much slower, and following a trajectory around the star a bit closer to that of a planet. Hence it will appear to wander in the night sky, but in a predictable manner.

Once it has decelerated into the inner solar system, it has slowed down to interplanetary speeds. At this point, its programming changes, and it begins a survey, flying from planet to planet on brachistochrone trajectories. It would appear to move between other planets and asteroids, and although it's trajectory would be predictable from night to night, where it goes after each stop might not be.

Perhaps during this phase, it journeys to a gas giant, and deploys cloud-scooping drones to refill its propellant tanks.

After it has finished its survey, it begins to accelerate again out of the solar system, becoming dimmer and slower, eventually appearing as just another star in the sky.

There isn't much of a reason for it to repeat this in a cycle. However, suppose there was a software bug that had escaped testing and QA. The probe inadvertently wipes its memory of the survey data, and loses track of it's position. It could decelerate and return to the same star system, thinking it has arrived at its next survey destination, repeating the cycle again and again.

Answer 2

From what you describe : not really

You are asking a 'thing' to act like a distant star and then like an object from the system. Those two cannot be compatible naturally as they do not have at all the same referential. A distant star is fixed from your POV. You cannot see its movement, you only see something moving because your planet is rotating on itself. For an object in your planetary system you have the rotating aspect but also the proper movement of the object compared to the star it is orbiting.

I can propose you something less erratic but that could appear to be a bit random :Two planets orbiting each other with one that has a near-to-zero albedo than could obstruct the light from the other one. From the POV of you planet the light would disappear each time the planet is obstructed. So you could have nights with no lights from the planet, sometimes eclipses (but as it would be a dot it would only disappear at once -almost-). If you set those two planets in another orbital plan and orbiting each other in this plan it would appear to be a bit random from a non-scientific perspective...

To be a bit more precise, if I chose a near-to-zero albedo for the secondary planet is only to make it seems like there is only one planet for a human looking at the sky (with a good telescope).

EDIT :As mentionned by another comment, something artificial could do the work, I only focused on the natural aspect.