SIGMA is a new flexible tool, inherited from DIANA opacity code (Woitke et al. 2016, Min et al. 2016), to compute dust properties of aggregates made of a mixture of distinct materials, including vacuum and ice mantles. Our goal was to deliver a code able to reproduce classic models from the literature on a very fast computation time, but also to deviate from them. Hence, SIGMA is freely available to the community. We rely on effective medium theory to compute effective refractive index and Mie theory applied to a distribution of hollow spheres to mimic non–spherical dust shapes. This method successfully reproduces exact results of other methods (like discrete dipole approximation) when selecting carefully the parameters related to dust shape. The core of dust grains is built using Bruggeman rule, while ice mantles are added as surrounding material thanks to Maxwell-Garnett rule. The last up-to-date version of SIGMA can be found on GitHub: https://github.com/charlenelefevre/SIGMA