Type 1 study results not calculating or not good

Hello all. I was doing a Type 1 study on a master ring at 94.996mm on a Mitutoyo snap gauge fixture and out of the 25 parts, the first time I had 0 variation. Everything read right on the money even removing and remeasuring 25 times. The second time I did it, I got 1µm of variation across the 25 readings, and it would bounce around between the two. For some reason, if I use 6 StDev and 20% for calculating Cg in Minitab, it is saying that the repeatability is 25.93% and the Cgk is 0.77 (Cg 1.31). The dimension on the actual production part is 94.985 ± 0.010mm. Am I missing something? With only 1 micron of variation, should this not be a great gauge? I get this feeling that I am just not putting a reading properly.