Return structural metadata for a document: filing type, source document type, total character count, named sections with character counts, and a recommended_call hint that guides whether to fetch the whole document or a specific section next. It also reports whether the document is mined (its datapoints have been extracted) and the ids of the published fundamentals mined from it, which the MCP uses to meter document reads. The endpoint never returns the document body. Documents whose filing type is outside the allowlist return 200 with recommended_call set to not_supported and reason populated. Earnings call transcripts are excluded from whole-document access and return 403; use /documents/keyword-search to query transcript content.