Project: Build a small tool to compute a Flow Alignment Score (FAS) from a 3D/4D vector field using three landmarks (outflow, apex, inflow). Input is my dataset with a point-data vector array (e.g., “Velocity”) plus three landmark points (provided as ParaView Point Sources or as coordinates). I can also run external Python or MATLAB scripts and provide ParaView saved states (.pvsm). Task: Derive three unit reference directions from the landmarks; for each point, compute a cosine-based score that rewards alignment with outflow, penalizes alignment with inflow, and mildly penalizes apex-ward direction; then map the result to a bounded 1–10 scale where 10 = best alignment. Output a point-data scalar named “FAS” that I can visualize and aggregate in ParaView. Requirements: Accept landmarks either as selected ParaView sources, by name lookup, or via manual coordinate entry; allow me to set the vector array name and the three weights; work with single/multiblock and time series; produce clear errors if inputs are missing. Deliverables: Either a ParaView programmable filter or lightweight plugin; or an external Python/MATLAB script that loads a .pvsm, computes FAS, and writes results (VTU/VTM) containing the FAS array; include a brief readme and a tiny synthetic test showing FAS ~10 for outflow-aligned vectors and ~1 for inflow-aligned vectors. Acceptance: Must run on my Windows ParaView install without code edits other than setting parameters; stable behavior with no crashes.