in/out stability circles enclose smith chart, but k-stability factor is all above 1?

ok, I created in/out single stub match microwave pre-amplifier. I want to know why my final result for in/out stability circles enclose the smith chart while my k-stability factor for all frequencies is above unity? Doesn't it mean, if stability circles enclose smith chart mean the circuit is unstable? If so, then isn't it counterintuitive to see that the k-stability factor is above unity for all frequencies, since you know that a circuit must be unconditionally stable, if k > 1? The stability k-factor, stability circles, and s-parameter responses for the amplifier are shown in the image below. Before I added all of my matching networks, the stability circles were right outside the smith chart, but when I found the final response, It also came out that the stability circles were enclosing the smith chart, is this okay, or not?

I would trust more the k-factor when is about stability.
Stability circles are printed on top of the Smith chart by using pure math equations. I don't know how your Matlab script is doing this. May be in a wrong way.

Are you sure your circles are stable outside?

There are two cases with stability circles
1 - stable inside
2 - stable outside

Rstab - stability circle radius
Dstab - distance between center of Smith chart and stability circle center

Four cases:

If center of Smith chart is unstable, and Rstab>Dstab, then circle stable outside
If center of Smith chart is stable, and Rstab>Dstab, then circle is stable inside

If center of Smith chart is unstable, and Rstab<Dstab, then circle stable inside
If center of Smith chart is stable, and Rstab<Dstab, then circle is stable outside

You may provide S11 S21 S12 S22 in complex or polar form and let me check

p.s. recommended: https://www.n5dux.com/ham/files/pdf/Avoiding RF Oscillation.pdf