They have near-fully overlapping meanings (I would be going out on a limb to say fully equivalent translations) with both the Arabic and German words having their primary use in expressing the meaning in "geometric shape" and also being analogous to the word form as in Platonic forms in their respective languages.
Supposedly stellen, the original verb form of gestalt, descends from the Proto-Indo-European *stel, which puts it at one t <-> k mutation away from the Arabic š-k-l (Or a shared ancestor with Hebrew š-k-l-l שכלל) based on examples like the palatalization in Modern Arabic dialects or Egyptian as an example contemporary to PIE if Loprieno p.31 can be cited:
[Proto-]Afroas[iatic] velar plosives *k, *g and 'k display two outcomes [...] either they are maintained as k [...] or they are palatalized into t, /c/ [...]
Also, mostly on pure fancy, I am interpreting the meanings conveyed in the cited Ancient Greek and Slavic descendants from *stel as containing the common conceptual element in grouping, bringing together, forming up etc. which can be a potential (even if very weak) indicator for *stel having a meaning related to form.
So this is where the too good to be true principle kicks in for me to look for factors that make it implausible for these two to have a common origin. What are some venues to exhaust for evidence on the contrary before actually pursuing this claim further?