This allows for $H \gamma \gamma$ and $Hgg$ to be effective tree level calculations instead of loop diagrams. This can simplify multi-loop calculations, and allows these processes to be put into tree level only software.
Simple Legrangian terms are1
$$\mathcal{L}_{H \gamma \gamma} = \frac{e^2_q \alpha}{2 \pi} \left( \sqrt{2} G_F \right)^{1/2}\left[ 1 - \frac{\alpha_s}{\pi} \right] \times F_{\mu \nu} F_{\mu \nu} H$$
$$\mathcal{L}_{H g g} = \frac{\alpha_s}{12 \pi}(\sqrt{2} G_F)^{1/2}\left[ 1 + \frac{11 \alpha_s}{4 \pi} \right] \times G_{\mu \nu}^a G_{\mu \nu}^a H$$
Using LanHEP, we can find Feynman vertex factors for the terms. First, the factor for the $\gamma \gamma H$ vertex:
$$-2 \frac{e_q \alpha}{\pi^2} (\sqrt{2} G_F)^{1/2} ( \pi p_1^\rho p_2^\rho g^{\mu \nu} - \pi p_1^\nu p_2^\mu - \alpha_s p_1^\rho p_2^\rho g^{\mu \nu} + \alpha_s p_1^\nu p_2^\mu )$$
and then the factor for the $ggH$ vertex:
$$-\frac{\alpha_s}{12 \pi^2} (\sqrt{2} G_F)^{1/2} \times \big( 4 \pi p_1^\rho p_2^\rho g^{\mu \nu} - 4 \pi p_1^\nu p_2^\mu -11 \alpha_s p_1^\rho p_2^\rho g^{\mu \nu} + 11 \alpha_s p_1^\nu p_2^\mu \big)$$
These vertices need to be added in CalcHEP:
G |G |h | |-HFA*alfSMZ/(12*pi^2) |4*pi*p1.p2*m1.m2-4*pi*m2.p1*m1.p2-11*alfSMZ*(p1.p2*m1.m2-m2.p1*m1.p2)
A |A |h | |-2*EQ*HFA*alfEMZ/pi^2 |pi*p1.p2*m1.m2-pi*m2.p1*m1.p2-alfSMZ*p1.p2*m1.m2+alfSMZ*m2.p1*m1.p2
These parameters:
HFA |0.0040935657 |sqrt(sqrt(2)*GF)
and these constraints:
pi |acos(-1) %Pi
EQ |(2/3)^2 %Charge of Heaviest Quark (For HEFT)
M. Spira, A. Djouadi, D. Graudenz, and P. Zerwas, Nucl.Phys. B453, 17 (1995), hep-ph/9504378 ↩