Scitex Types

Type aliases and runtime type guards for scientific Python — unions across NumPy/pandas/torch arrays, matplotlib color inputs, and list-of-T predicates. Public API (5 symbols) — `ArrayLike` (union alias spanning `list`, `tuple`, `np.ndarray`, `pd.DataFrame`, `pd.Series`, `xr.DataArray`, `torch.Tensor` for type hints), `ColorLike` (union alias for matplotlib-compatible color inputs — RGB tuples, hex strings, named colors), `is_array_like(obj)` (runtime predicate for array-ish inputs), `is_list_of_type(obj, T)` (runtime check that every element is of type T), `is_listed_X(obj, X)` (alias for is_list_of_type). No CLI, no MCP tools. Drop-in replacement for `numpy.typing.ArrayLike` (narrower — NumPy only), `typing.Union[np.ndarray, torch.Tensor, pd.DataFrame, ...]` hand-rolled unions, `isinstance(x, (list, np.ndarray, ...))` chains scattered across code, and `all(isinstance(el, T) for el in lst)` one-liners. Use whenever the user asks to "type-hint a function that accepts ndarray/DataFrame/Tensor", "check at runtime whether something is array-like", "validate that a list contains only floats/ints/strings", "type a matplotlib color parameter", or mentions `ArrayLike`, `ColorLike`, `scitex.types`, duck-typing array inputs.

mkdir -p ~/.claude/skills/scitex-types
curl -L https://claudskills.com/skills/scitex-types/SKILL.md \
  -o ~/.claude/skills/scitex-types/SKILL.md