The job guarantee is pretty well encapsulated in the discussion of inflation, since MMT focuses on NAIBER rather than NAIRU to focus inflation targets.
(Wikipedia) MMT's main tenets are that a government that issues its own fiat money:
1. Can pay for goods, services, and financial assets without a need to first collect money in the form of taxes or debt issuance in advance of such purchases;
The intended thrust of this principle is that you can spend first and tax later because the monetary system is not about accumulation but about equilibrium, and the equilibrium isn't about total balance but about interest rate and employment. The first part is the assumption: fiat money will be able to pay for goods, services, and financial assets.
This need not be true, and particularly tends not to be true at the beginning and end of fiat currencies. At the beginning of a currency, the issuing entity has to establish faith in the system. And as a currency fails through hyperinflation, it ceases to have any meaningful buying power. The point where you call it failure is obviously a point for debate, but I would suggest that your currency has proceeded past the point of failure if, when transacting for normal daily purchases (food, transportation, small durable goods), money is exchanged not by reading face values but by weighing large quantities of notes or similar (e.g. bread costs two bundles of orange notes.)
2. Cannot be forced to default on debt denominated in its own currency;
The only situation where this ceases is to be true is one where the issuing government does not possess the resources to do the requisite printing or, probably even more unlikely, to declare new money supply digitally. For all intents and purposes, this statement is true, but again it doesn't acknowledge the fact that all lending to a hyperinflationary government will cease when there is no faith in the value of that currency. This drying up of credit has the same effect as default, particularly since the debt was never secured by collateral.
Nonetheless, the effect will be similar to defaulting on secured debts, since so little of an economy would be functioning at this point that the hyperinflationary state would require imports which could only be acquired through selling off of hard assets and land.
3. Is limited in its money creation and purchases only by inflation, which accelerates once the real resources (labour, capital and natural resources) of the economy are utilized at full employment;
This part is trickier to handle. We can probably agree that inflation means the widespread rise of prices across goods and services (though the measurement is a tricky issue that we should put aside for now.) Full employment is the sticking point, because MMT replaces NAIRU with NAIBER. NAIRU is a point of uncommon humility for economists, because it acknowledges that, if we push for zero unemployment, we get accelerating inflation. NAIBER, on the other hand, pushes to extinguish unemployment through job guarantees in the form of government jobs (the 'rate of unemployment' of NAIRU having been replaced by 'buffer employment ratio' in NAIBER.) So, why wouldn't this be inflationary just like in the other scenario?
Well, the NAIBER principle suggests that you pull people away from the inflationary sectors of the economy into lower-paid government jobs. In this way, there's less inflation because wages aren't being allowed to shoot up in booming sectors of the economy. The methodology for convincing people to make job changes away from their chosen work to something new and lower paid is a question I haven't seen answered anywhere, but I doubt that the changes would be effected peacefully or voluntarily.
In the meantime, all this government payroll for the lower-paid jobs will result in more inflation from the extra government spending required.
4. Recommends strengthening automatic stabilisers to control demand-pull inflation[10] rather than relying upon discretionary tax changes;
This statement just means that there should be a set of policies that can be set long-term which would be preferable to tax changes with the changing political winds. Every other political wind that comes along also agrees with this statement and, like MMT proponents, they think that their ideas are the right one.
This is basically arguing for the rule of law. The devil is in the details, but diving into those would require us coming to a common understanding of those details first. I'll leave off there on this point.
5. Bond issues are a monetary policy device, not a funding device.
This formalizes an underlying assumption of the modern US federal government that we can always borrow more money. That clearly works at present, but it is more prone to failure than MMT proponents acknowledge. If the government doesn't issue bonds to attain funding for (or to counterbalance, per MMT framing) spending, the currency is debased. There comes a time when that accelerates painfully and it becomes worth the world's efforts to disentangle from the USD.
__________________________
MMT is a theory that starts with the premise that there will be no catastrophic end times, but working under that premise is the surest way to bring them about more quickly. MMT has a long runway to failure in the US (if it is done slowly) because there is so much financial infrastructure denominated in USD, and there is a lot of underlying value in the US. There are much slower ways for a government to fail, and those are the ones I want. Businesses should have some acceptance of downside risk, because a failed business is only pretty bad. A failed country is a whole different level of catastrophe, and our governments should be accordingly allergic to government-ending levels of downside risk.
Edited for formatting.