I don't think that even LEO Spy Satellites have that resolving power. Of course, that information would be classified. Does any HN reader know what a reasonable guess at their resolving power would be? I would love to know.
Commercial satellite imagery is available at 0.25m resolution these days, so assuming an order of magnitude better that gives around 2.5cm resolving power, or one inch. I think I've heard about 5cm quoted somewhere, as well. So, maybe just good enough to detect the difference between small-arms, and certainly enough to read identification markings on an aircraft, or to determine the name of the newspaper, but not to read its headlines.
I wonder what the real limit is for resolution, due to the distortion of the atmosphere.
I guess that by taking multiple photos in quick succession from slightly different points you could reconstruct a clearer image. It's not my field at all though, so I may just be randomly guessing.