SpecKit: An integrated toolkit for neutron spectrum unfolding using activation reactions

In response to the need for neutron spectrum unfolding based on activation reactions, this study presents SpecKit , a reproducible and user-friendly software workflow. The method models the problem as a linear system using an activation response matrix M A and an observed activity vector b A , and estimates the unknown flux vector through residual minimization combined with log-smoothness regularization. Uncertainty quantification is achieved via repeated fitting and Monte Carlo sampling, with resulting group-wise uncertainty bands provided. The software includes modules for cross-section and matrix preparation, group flux unfolding with uncertainty estimation, and results visualization, all integrated within a graphical user interface (GUI). Key output metrics include group-wise flux, total flux, relative error. Performance evaluation is carried out using MCNP -generated synthetic scenarios under systematically designed prior-group flux mismatches, with analysis of deviation across thermal, epithermal, and fast neutron energy regions. Results demonstrate that the proposed method consistently corrects prior bias, maintains low deviation across energy ranges, and provides well-calibrated uncertainty estimates under varying levels of prior shift and measurement noise. A minimal data package and reproduction scripts are released alongside the project to facilitate community validation and further development.