question: Arabic morphology

Okay, I'd been wondering about this for a while but it's taken 2 hrs.
with Mark Rosenfelder's neat but not-useful-to-me-at-the-moment Sound
Change Applier for me to be able to articulate this.

Arabic uses three-consonant verbs, correct?  I borrowed this for
Chevraqis (which is a compound, I haven't figured out for what yet, and
the name is subject to change anyway).

I was trying to use the SAC and ran into a problem.
How does the morphology evolve?  Right
now I have something like this (Aragis -> Chevraqis):

imperative: aCCaCu -> CiCaCu
causative: CiCoCu -> CiCoCu
generic: CaCaCu -> CeCaCu

etc.

Now that I look at essays on how sound change happens, a wholesale shift
like this seems unlikely.  My question is, as far as this morphology
business goes, do I apply sound change across a pattern (what I have
above) or per individual word?  As it stands I can't figure out a way to
get the SAC to work for me at all, though I'll probably be able to use it
on irregular things and pronouns and so on.  I got those patterns by
applying rules (pretty simplistic ones, I'm afraid, but I haven't found
much on common vowel shifts on the web, and a friend has my _Historical
Linguistics_ by Crowley) to the vowels, and then by mutating the
consonants according to *those* rules.

The SAC doesn't let me apply pattern-rules like the above, though; maybe
you can set up the variables in such a way, but I can't figure it out, or
perhaps you could edit the source code, but my C is nonexistent (and my
Java's awful rusty--the closest programming language I know).  I tried
applying the sound changes *individually*--the program goes down the list
sequentially--and what came out was considerably different than what I
got using my tedious, by-hand method.

Am I missing something here?  Is there a good source on Semitic
languages?  I scoured the web and came up with nothing (in English,
anyway).  Help!

YHL