@squidcloud/client
    Preparing search index...

    Interface DocumentMetadata

    Document-level properties read from the source container during extraction. Every field is optional and best-effort: when a container does not expose a property (or parsing fails) the field is simply absent. Consumed by schema-driven metadata extraction as the deterministic source for matching knowledge-base metadataFields before falling back to LLM extraction.

    interface DocumentMetadata {
        docAuthor?: string;
        docCreatedAt?: string;
        docModifiedAt?: string;
        docTitle?: string;
        docType?: string;
    }
    Index

    Properties

    docAuthor?: string

    Document author (e.g. OOXML dc:creator, PDF /Author, front-matter / <meta author>).

    docCreatedAt?: string

    Creation timestamp, ISO 8601.

    docModifiedAt?: string

    Last-modified timestamp, ISO 8601.

    docTitle?: string

    Document title (e.g. OOXML dc:title, PDF /Title, front-matter title, <title>).

    docType?: string

    Source document type: pdf | docx | pptx | xlsx | md | html | txt.