I agree it looks like it's clipping, but the OP279 has rail to rail input and you only seem to be using fairly small signals, so it should not be clipping.
Are you using 10x probes, or 1x probes, or just BNC cables, to the oscilloscope? Is the oscilloscope input impedance 1M, or 50 Ohms?
I just using 1x probes and the oscilloscope input impedance is 1M.
On 2nd review, I noticed your first stage DC oscillograph has 2V/DIV and shows almost 5V of DC. I think that is your problem. Did you set up your signal generator as 2.25V, but then not terminate it with 50 Ohms? Most generators assume 50 Ohm termination, so without termination the signal is 2x too large.
I suspect that the first stage of dc has been amplified twice, and that the first stage amplifies the output close to the supply voltage when there is no signal the first stage input
the first stage output