Get the tree structure of an indexed data source.
Works with all source types:
Documentation: Returns a filesystem-like tree structure of indexed pages, making it easy to browse documentation without remembering URLs.
Research Papers: Returns the document structure with sections and subsections.
HuggingFace Datasets: Returns a tree showing dataset splits and columns, with metadata about row counts and sampling strategy.
API key must be provided in the Authorization header
Flexible data source identifier. Can be:
Tree structure retrieved successfully
Nested tree structure of documentation pages
Formatted text representation of the tree
Type of tree structure returned
documentation, huggingface_dataset Base URL of the documentation (documentation sources only)
Total number of indexed pages (documentation sources only)
Dataset identifier (HuggingFace datasets only)
Dataset owner/organization (HuggingFace datasets only)
Dataset description (HuggingFace datasets only)
Available dataset splits (HuggingFace datasets only)
Dataset column names (HuggingFace datasets only)
Total number of rows in dataset (HuggingFace datasets only)
Number of rows actually indexed (HuggingFace datasets only)
Sampling strategy used (full or sampled) (HuggingFace datasets only)