cascode switching

I've built circuts sort of like this. It didn't involve points. Several things can be going on and many of them are bad. I'll take them sort of in order.

At the instant of the points closing, the fact that the coil has a largish stray capacitance makes a current spike. Check that the 10R limits the base current enough to keep the transistor safe.

At the instant the points open the transistor is saturated. It will take it a little time to stop conducting. During that time, the emitter goes above 12V and the E-B junction can break down.

Just as the transistor is actually turning off, there is a large collector bas voltage and a collector current flowing. This can lead to 2nd break down issues.

When (and if) the spark happens on the secondary, a burst of high frequency stuff is produced on the primary. If the collector to emitter capacitance is enough, this can also breakdown the e-b junction. I don't think the auto-ignition coil make a high enough frequency for this to happen.

When the primary rebounds, it will likely swing well below the +12V. This will forwards bias the collector junction and flow a current. Check that the 10R is enough to keep that current within bounds.

The same rebound caused current will pump carriers into the base so that when the collector flies up again, there is another recovery time.

I suggest you try:

Place a diode in series with the collector so that the current can't reverse.

Add a Baker clamp to the transistor.

Hook a string of 4 diodes in series from the transistors emitter to the