BRAINS (BLR Reverberation-mapping Analysis In AGNs with Nested Sampling) dynamically models the broad-line regions of active galactic nuclei using reverberation-mapping and spectro-astrometric observations. It couples flexible geometric and kinematic BLR models with radiative transfer and line-response prescriptions to reproduce observed emission-line light curves, spectra, and spatial signatures. The code employs nested sampling to infer BLR structure and black hole mass, providing posterior distributions for physical and nuisance parameters and enabling rigorous model comparison. Implemented in C and Python, BRAINS supports configurable data sets, model components, and priors, and includes utilities for data preparation, parameter estimation, and visualization of inferred BLR and black hole properties.

RECON measures power spectra and reconstructs time series in active galactic nuclei (AGNs) by modeling their stochastic variability in the frequency domain. The method treats the Fourier transform of AGN variability as a set of complex Gaussian random variables, parameterizes the resulting stochastic process, and transforms it back into the time domain to fit observational data. Model parameters and their uncertainties are estimated within a Bayesian framework.

PyCALI intercalibrates astronomical light curves using a Bayesian MCMC framework. It applies additive and multiplicative factors to light curves to bring them into a common scale by modeling the variability with a damped random walk process. Systematic error factors and error scale factors can also be incorporated.

PIXON analyzes reverberation mapping in active galactic nuclei. The pixon method uses pixons (instead of pixels) as the basic unit, which are able to adjust pixon sizes to achieve locally optimal resolutions according to the information content provided by the data. Within a pixon, the pixon algorithm smooths the solutions as much as the data allow. The terminated criterion of the pixon method is to find the fewest number of pixons that still adequately fit the data. As such, the pixon method optimizes solutions not only by testing the goodness of fit but also by reducing the complexity to be optimal. The issue of overresolution or underresolution of the solutions is also significantly alleviated.