Space based weapons are extremely limited in practice. The problem is that at any given time 90% or more of your orbital assets are over the pacific ocean, or the arctic, or antarctic, or anywhere other than where you want them to be and you can't change that very easily.
I'd love to know if they've ever been tested in reality. If succesful (especially as a first strike), you could theoretically launch from a high apogee using cold gas to adjust your heading, and have plausible deniability.
A LEO sat at 300 km altitude does a full orbit every ~90 minutes.
Put a few hundred missiles in orbit, give them enough spare fuel to travel a decent distance on their own, and you have a pretty substantial worldwide strike capability.
Do you know what the difference is between an ICBM and a warhead predeployed in orbit?
The ICBM can hit anywhere (instead of just where the orbit track happens to pass), in less time (or at most the same time, since deorbiting takes around half an orbit) with a better mass fraction (since it doesn't have to reach orbit and then deorbit afterwards) while being more accurate (you know exactly where the launch platform is, while satellites are harder to locate) and less vulnerable (you can harden an ICBM silo far more effectively than a satellite).
Really, it's hard to think of something that satellite anti-ground weaponry does better.
There is a lot of icbm detection. Unplanned rocket launches could lead to global nuclear annihilation. What happens if something seems to deorbit above the US? How much early warning and procedure is there?
Especially if you want to deliver conventional munitions to a non-nuclear power, a system like this seems better than an ICBM.
Except that detectability is an advantage, in the doesn't-set-off-global-thermonuclear-war sense. If your nuclear (or command) assets can be removed at any time without enough warning to launch, that means you have to launch now before that happens.
If you want to hit something with a conventional warhead in a non-peer state without warning, why not use a stealthy cruise missile instead?
This handwaves a lot of stuff. The least of which is that with the way satellites orbit, even though they orbit in 90, they’re not over the same ground track again for like 12 days. It’s not the case that a weapon used satellite would have a firing solution every 90 minutes.
Smallest nuclear bomb (not dirty bomb) made was 23kg, literally suitcase size. However if you have one and want to use it, you're probably better off sending it in on a shipping container or truck (or commercial or private plane)
Classified payload. Ostensibly failed to detach from the payload adapter, as claimed by anonymous sources... Lot's of speculation that this was a media charade to sow confusion about a stealth sat.
"That launch deposited a payload into geosynchronous orbit but, given the stealth/deception hypothesis, there remains the possibility of other, undetected payloads"
There remains a "possibility" that dinosaurs aren't extinct and they're hiding somewhere. Just because civilian astronomers haven't found such payloads doesn't mean those payloads actually exist. And hypothetically even if those payloads do exist it doesn't mean other nations haven't tracked them.
Wikipedia isn't a reliable source for this stuff. Look at the basic physics involved.
While amateurs can do an amazing job of detecting satellites, I think it is pretty likely that there are smaller satellites which are operating in extreme low power mode, waiting until a conflict where they are required. Whether there are enough to replace GPS and comms lost due to anti-satellite attacks and resulting debris is a bigger question.
It isn't physically possible to get useful GPS service from a small satellite. This is just basic physics. Calculate the power required to transmit the necessary signals. You need some fairly sizable solar panels. No way to hide those.