Changes in version 0.1.1                        

  - Added automatic detection and joining of split Seurat v5 Assay5
    layers such as counts.sample1, counts.sample2, and matching data.*
    layers.
  - Counts-only Seurat objects can now be written to AnnData without
    normalization; counts are placed in X and layers["counts"] with an
    explicit conversion note.
  - inspect_sc() now reports detected count and normalized-data layer
    groups for Assay5 objects.
  - check_input_info() is retained as a compatibility wrapper but points
    users to inspect_sc().
  - Improved assay-cell, feature-metadata, and reduction alignment
    during Seurat extraction.

                 Changes in version 0.1.0 (2026-09-29)                  

  - Initial development release.
  - Adds a common convert_sc() interface for Seurat,
    SingleCellExperiment, AnnData, and Loom.
  - Preserves aligned counts, normalized expression, cell/feature
    metadata, and compatible dimensional reductions.
  - Adds Seurat classic Assay <-> Assay5 conversion helpers.
  - Uses reticulate::py_require() instead of a hard-coded Conda
    environment.