Don't turn the breaker on again at this time. Every time you do so damages the circuit conductor more. You probably need an electrician.
Here is what I would do. Turn off, unplug everything on that circuit. Remove the wire from that breaker in the panel. Using a "pigtail" lamp socket, wire nut the white wire of the socket to the wire removed from the breaker. Put the black wire of the socket to the breaker. Install a lamp (incorrectly called "bulb"), 15 watt, 25, 40, 60, whatever, into the socket. Connect an insulated wire long enough to reach any part of any of the cabinets, to a ground, either in the panel or at a receptacle on an unaffected circuit near the cabinets. Get a second pigtail socket and wire nut one of the leads of it to the ground wire. Install an identical lamp in that socket. CAUTION-- you will soon be handling 120 volts. Turn on the breaker and the lamp near the breaker should light. Then touch the unattached lead of the other socket to the screws/nails that the installers drove into the wall, starting with the base cabinets. If the lamp lights (probably dimly), that is the offending fastener. The lamp at the panel permits putting 120 volts on the circuit while also acting as a current limiter.
If a live fastener is found, the cable into which it is driven it MUST be replaced. Do whatever is required to accomplish that. Simply removing that fastener may seem to solve the problem, but the conductors in the cable may be damaged and reduced in area which will create a hot spot and a future fire in the wall.
Not high tech, but foolproof trouble shooting.

Thanks for your reply, but I have no idea what you just said -:)

I do agree that we need an electrician, but my question is this -- is there anything the cabinet installers could have done to cause this problem?

They were using a big power drill, could that have tripped something (all working outlets in the kitchen are fine, but the overhead light does not
The countertop outlets are wired for, but not in yet. So I don't have a GFI outlet yet.

I think something must have happened during their work here today, yet I don't have enough knowledge to come up with a reasonable explanation.

In any case, the breaker tripped and won't reset.

They put a screw through a cable in the wall while mounting your new cabinets. This is not necessarily their fault. There should have been nail plates/guards in place to prevent this.

Depends how long their screws are. The nail plates are only required if the wire is closer than 1 1/4 from the face of the stud.

It's not the installer's fault if the original electricians ran the wire and didn't protect it properly. Call in an electrician to hunt down the culprit screw. It will take a bit of time, so be prepared for a not inexpensive service call. The cabinet will need to be pulled and the wire re re run and then you'll have some drywall patching as well.

"It's not the installer's fault if the original electricians ran the wire and didn't protect it properly"

The NEC ONLY requires a screw plate if the edge of the hole is clser than 1.25 inches to the face of the stud.

It is for protection from DRYWALL screws, not the longer screws used to hang cabinets.

It is THEiR fault even though it was likely accidental.
The should have used a circuit detector or metal detector to check for cables.

At least it caused a short.

I have seen plenty of 'hot' screws found years after the fact.
If the screw only hits the hot conductor nothing may happen at all, except to a person who manages to touch the crew head and electrical ground at the same time.

I have to agree, sounds like a screw penetrated a wire. I like Bus Driver's method, I hadn't heard of that before, but was thinking a non-contact circuit tester might do the same thing - the ones where you press the button and hold it near a live wire or outlet, and it emits a tone. It should be really loud at the offending screw.

That's kind of an interesting conundrum. From a point of holding your cabinets up, the longer the screw, the better. The disadvantage of that we can see in this instance!

Next job I do in a new construction place, I'll have to look and see what the sparkies do in kitchens - it would make sense to mandate nail guards in any kitchen wall likely to have cabinets or shelving.

It'd also be wise not to run wires at obvious cabinet mounting points. Even nail guards are not foolproof, a sharp screw would go through with some effort.

The fault will have to be traced and the wire re-run. Perhaps a junction box could be put in the cabinet where the wire was damaged and the wire spliced. Not ideal, but less disruptive.

OP, it's not the drill doing something to the breaker, it was what they were driving. Don't waste any time on this.

Are cabinet installers expected to check for cables as a matter of course before every install? I would think a non-contact tester wouldn't detect it (wire depth) And a metal detector? What sort?

"Perhaps a junction box could be put in the cabinet where the wire was damaged and the wire spliced."

Very unlikely.

You would need more than an extra foot of wire that could be pulled into the new junction box to meet the wire length rules.

Generally it requires two junction boxes to get enough wire to repair a damaged section.
The boxes must remain accessible after the repair (and the cabinet install) is completed.

"Are cabinet installers expected to check for cables as a matter of course before every install? "

Unless they want to pay for repairs to wiring it would be in their best interest.

I would bet this one will from now on after he gets the bill.

Non-contact sensors are not all the sensitive.

Many electronic 'stud finders' are much better at identifying electrical cables.
There are also even more sensitive detectors used for identifying circuits.
They place a signal on the cable (often by plugin them into a live circuit but sometimes by just hooking them to a wire (those use batteries)) and then have a receiver that detects the signal (often with adjustable sensitivity to let you 'zoom in' to a spot.

I have a transmitter/receiver wire tracer - I bought it mostly to figure out which outlet is on which breaker. It's pretty cool and can definitely show the wire within the wall, but it does have its limitations. The biggie is you need to plug it into an outlet or light socket, and trace it back from there, which means you kind of have to know which outlet to check, and you never know what's being run through there - it could be wires going to another room.

There's probably a little bleed-through on all lines if they're on the same 120v leg...maybe even both legs, since the neutrals are tied together.

OF COURSE any junction box will depend on there being enough slack. It will also mean that the only wire needing to be replaced will be that from the break to the loads, not all the way back to the meter.

"Perhaps a junction box could be put in the cabinet where the wire was damaged and the wire spliced. Not ideal, but less disruptive."

You plainly said "a junction box," and since there is rarely adequate slack in a wiring run (especially if the sable was not fished into a void) TWO boxes are always required.

You need wire length from the face of the junction boxes to meet code, adding significant length to the 'slack' that would be required.

Geez, you really are a pedant. Two boxes, or six, or whatever you say it takes, then.

Actually, in this situation, one box might well do it. Assuming they're ripping out a fair bit of drywall, they do enough to get that slack required, then run new wire to a second one if needed.

"Geez, you really are a pedant."

And you are just plain ignorant.

If you do not really understand what is needed, why say anything?

You used to be really helpful, now, most of the time you're just nasty. What's up with that?

Bus driver. That was a very intriguing method for finding the problem screw. What if the screw were shorting all 3 wires together when it penetrated which is likely possible since the breaker trips? When she put the first lamp in and turned the breaker on, wouldn't that just burn the lamp out near the panel? I am not trying to be adversarial. I just like tossing this stuff around and wondered what could possibly happen. I guess removing the ground and the neutral from the buses in the panel from the problem circuit might work as long as there were no other common grounds (or neutrals-not that they should have been) tied together with the problem circuit like maybe in a multi-gang switch box or something where 2 circuits' grounds might have been tied together.

There are many options to be considered when troubleshooting. In this case a digital voltmeter could have have one lead attached to a suitable ground and the other lead probe touching all the cabinet wall fasteners in turn. Try it on volts and then resistance. Obviously at least one of the fasteners will likely give a reading unlike the others just from ghost voltage and different resistance. It doesn't matter much anyway because new wiring is needed.

Analogue voltmeters are often better for this sort of thing, digital ones are prone to showing phantom voltages.

For that matter, a neon tester would probably work too.

Ah, no. You want to see a higher DVM phantom ac voltage only on the shorting fastener(s) in question in this case because the circuit cannot be powered up. If you also have an analogue VOM multimeter available to use for the fastener(s) resistance check to ground, that would be ideal.

" most of the time you're just nasty."

Some of the replies have become just plain ignorant.

The next set of questions that will appear is 'Why cannot I just use a single box to fix the damaged cable' when someone calls an electrician and he has to use two JBs.

And the answer is generally no without replacing some cable (and the other end of the cable is generaly fastened within a fw inches of the box).

A the fact you started off ignoring.

It takes two JBs to splice in a new cable section just about very time.

You can get away with one IF you are willing to replace one end of the damaged cable (with the damage to walls that typically requires) but generally it takes two JBs.