The simplest way would be with a DC-blocking cap, however you may
get unusably short pulses out at high revs or it may not discharge
between i/p pulses. My preferred method would be to buffer the Hall
effect sensor with a one-shot made from a couple of 4001 NOR gates.
That would give you a constant width pulse whatever the rpm (if you
set the pulse duration to slightly below that of the max rpm that
you expect). The one-shot will trigger only on a rising edge, which
makes it safe if the magnet happens to stop over the sensor. At 12V
100k and 10n should give you about 1ms, but gates do vary

Attachment converted: creation:oneshot.gif (GIFf/JVWR) (00015CC5)