You have to be very careful with your definitions here. Following the basic definitions (e.g. see use of terms in Mankiw Principles of Economics):
Value: Value depends on your marginal utility and typically it is the amount of money you are willing to give up for something.
Wealth: Is by the value of net assets.
Income: Is the net return to some activity.
You are right that economy is not necessary zero-sum game (although there are economic interactions which can be zero-sum), but your use of the term wealth is improper.
If someone pays you to fill the soda machine, you have costs \$0.5 but get paid \$1 your income will be \$0.5, you created at least \$0.5 of value because the person who paid you clearly valued the filled vending machine for at least \$1 otherwise they would not exchange that 1 dollar for your work, but at the same time you had 0.5 costs, so you are creating at least \$0.5 extra value.
The wealth is accumulated by saving (whether you save by buying house or putting money on your account value of your net assets increases). Hence if you have \$0.5 income and you save it all you also create \$0.5 wealth. If you consume half you only create \$0.25 wealth and so on.
So in the above example with the vending machine, does the federal reserve have to create an extra 50 cents of the currency in order to avoid deflation?
Here the answer is maybe. Inflation/deflation does not depend just on the value of output that you create.
Inflation/deflation is just positive/negative change in the price level which is in turn determined by the money market equilibrium. The money market equilibrium, in its simplest form is given by equation of exchange (See Mankiw Macroeconomics pp 87) as:
$$MV=PY$$
Where $M$ is the money supply, $V$ velocity of money, $P$ price level and $Y$ output.
Solving for price level and log-linearizing (so % changes in right hand side variables give us the % change in $P$) we get:
$$\ln P=\ln M + \ln V − \ln Y$$.
If you perform those extra services you increase real output by \$0.5, ceteris paribus, you are correct that Fed would have to create additional \$0.5 dollars to prevent inflation. But the, ceteris paribus, assumption might not hold in real life because velocity of money can change as well (plus in more complex models of money market equilibrium expectations play role and so on). Consequently, the correct answer here would be maybe. Under ceteris paribus assumption yes, but it is not guaranteed.
PS: Note sometimes even economists in casual speech equate wealth creation with value creation, so you might have heard some pundit or economists use similar example to argue wealth was created, but that is not how the two terms are rigorously defined in the literature.