MOTIVATION: Light chain amyloidosis is a disease where misfolded antibody light chains (LCs) form toxic amyloid fibrils, leading to organ damage. Although LC overproduction occurs in all cases, only certain individuals develop the disease, suggesting that specific LC sequences and properties drive amyloid formation. This process is complex, involving both protein sequence and environmental factors, but mutations that destabilize the LC fold are linked to amyloid aggregation. Despite the significance of the disease, our understanding of LC fibril formation remains limited due to the lack of extensive data and technical challenges in studying amyloid structures. To address this, a tool is needed to compare unknown LC sequences with known structures and predict which amyloids are likely to adopt new conformations, guiding experimental investigations. RESULTS: HMMSTUFF addresses this by using a Hidden Markov Model to generate similarity scores between LC sequences and existing PDB templates, eventually modeling the LC amyloid structures similar enough to known templates. HMMSTUFF on one side expands our understanding of LC amyloid fibril conformations, and on the other highlights the gaps in our current knowledge of LC structural space. AVAILABILITY AND IMPLEMENTATION: HMMSTUFF is available as pypi package and as source code at https://github.com/grogdrinker/hmmstuff.