Mine was purchased from an older mechanic, so why it works? I'm not sure. Maybe park tool sold the combo together at some time. The set-up looks like Randy's.

HA! very clever.

you may want to get a hard bolt or the tip will get messed up with use. metric bolts have a number on the head that indicates hardness. 10.9 is a nice hardness for tools and common in finer pitch threads. cylinder heads in engines often use oddball bolts, may pay off asking your local mechanic if he has some laying around (they get changed sometimes)

Yes, this jibes with my measurements on both the early (two-ended) Park tool and their later version with the thicker handle and with finer threads.

Both versions have 9/16" SAE-threaded driver, so can't fit the 12mmx1 Stronglight puller threading.

The thread pitch is 18TPI for the early Park puller and 24TPI for the later one.

A VAR 22/2 puller I have here uses a 12mm driver, but is has 1.25mm thread pitch.

What I found most interesting has to do with the VAR tool's listed "23.35mm" spigot threading, which I measure to a major diameter of only 23.01mm.
Compare that to the early Park tool's actual 21.89mm and there is 0.12mm difference, while the genuine Stronglight tool has a major diameter of only 23.13mm.

What I experienced in the distant past was that the early, 2-ended Park tool successfully pulled several Stronglight 93 cranksets from their spindles without stripping.

While my experience surely must fall somewhat to ignorant good luck, another, better approach is to use the early park tool, but with the handle only tightened to a modest, say half, fraction of the normal amount of tightening needed to free a crankarm.
If the "rider" then stands on the pedals and "hops" (with the tool still providing it's pulling force), then reverses the cranks 180-degrees and repeats this perhaps 4-5 times, the Park tool handle can then be tightened a bit more, and the hopping process repeated, again reversing the load 4-5 times.
Note that I am by no means heavy myself, so this should not require "violent" (or in any way destructive) levels of hopping force!
After several rounds of hopping and re-tightening of the tool's handle, the stuck crankarm will release with far, far less tension needed from the tool, this because the square end of the spindle twists slightly under load, with the surface friction retention eventually giving way to the twisting motions of the hopping (and simultaneous force of the puller's tension).
I've used this same technique on "impossibly"-stuck crankarms where the tool risked breakage (or where the tool handle actually broke off), and have never failed to remove such a stuck crankset with a more-modest level of force needed from the puller.

Out of curiosity, I also just now tried threading my "23.35mm" VAR 22/2 tool into an old TA crankarm, and even though the threaded spigot measures just 23.01mm, it would not thread into the TA crankarm, even though the only slightly-smaller Park tool easily does.

That would be clever on the part of that mechanic, using the park driver with integral handle for convenience in use (or to restore damaged threading).

The Stronglight puller would have only had to be drilled and tapped to a common 9/16x18tpi ("9/16 NF") for the early Park handle to fit.

This might be the best answer for the OP (pontius) if a Park tool is perhaps already at hand, AND if the OP has some machining experience (plus the drill press, drill bit, cutting fluid/oil, goggles and tap handle that will also be needed).

But I would prefer using the stronger, finer-threaded (24tpi) handle/driver from one of the later Park tools if I could also find a 9/16x24tpi ("9/16 NEF") tap without too much searching or expense. See below:

https://www.ebay.com/itm/9-16-24-NEF-...item2c9c1f25e5