I think I have it on some other drive, and will try to dig it up. However, it is a bit of an overkill for what you describe, and what you need shouldn't be terribly hard to write yourself (I'm assuming that you want to render the text across multiple lines in a bottom-expanding box of predefined width):
-Divide up the text into segments, one for each time you switch font (probably easiest to just pass a string array in the first place, and a corresponding array of FontTypes; alternatively you could write some text parsing functions that would allow you to put formatting tags into the string, but that would be a lot more work)
-Figure out how to render each string segment inside the defined rendering area (i.e. the given width). For each string segment:
--calculate its size when rendered in the appropriate font, using GetTextHeight()/GetTextWidth(). Keep track of these values.
--work line by line. Calculate and keep track of the starting x-position of each segment (by adding the widths of the preceding segments together), and the max text height seen on this line so far.
--given the starting x-position of the rendering and the width of the rendering area, determine if you need to line-break (doing line-breaking internally within a segment is probably the most complicated part, so avoid it if you can)
--once you've reached the end of the line, use the max text height (for line spacing) and the various x-positions to position and draw each string segment, and move on to the next line.
I don't remember how AGS DrawString() positions text vertically, but I suspect it's by the upper edge. So you may need to define a vertical displacement for each font so that they'll all have the same baseline.