On Saturday, October 14, 2017 at 6:36:57 PM UTC-7, bob prohaska wrote:
[about anti-flux]

It seems as if correction fluid might be a pretty good bet. Near as I
can tell the pigment is titanium dioxide...

the broken tab is only about 1/8" square. Can anybody suggest some sort
of putty that will resist brazing heat, hold the (very small) parts in
place and not sinter to something indestructible when heated? Maybe
glazier's putty, but that can get rather hard. Fire stop caulking?
modeling clay? Perhaps something made from household chemicals like
talc and water?

Modeling clay (the greasy kind) or putty are no good. Casting sand
and a little molasses is traditional for high heat, though. Or a mineral clay
might work, I know painting a little clay slip onto copper wire keeps hot
glass from sticking.