@squidcloud/client
    Preparing search index...

    Interface AiKnowledgeBaseFileContextRequest

    Request structure for adding file-based context to an AI agent.

    interface AiKnowledgeBaseFileContextRequest {
        contextId?: string;
        discardOriginalFile?: boolean;
        extractImages?: boolean;
        imageExtractionModel?: AiChatModelSelection;
        imageMinSizePixels?: number;
        metadata?: AiContextMetadata;
        options?: AiKnowledgeBaseContextFileOptions;
        preferredExtractionMethod?: DocumentExtractionMethod;
        scopes?: string[];
        type: "file";
    }

    Hierarchy (View Summary)

    Index
    contextId? discardOriginalFile? extractImages? imageExtractionModel? imageMinSizePixels? metadata? options? preferredExtractionMethod? scopes? type
    contextId?: string

    The id of the context. Optional: when omitted, the server generates one and rejects the upload if another context in the knowledge base already holds identical content (content-level dedup). Provide an explicit id to address a specific context — that overwrites it unconditionally and allows duplicate content across different ids.

    discardOriginalFile?: boolean

    Whether Squid discards the downloadable original once the file is extracted; the extracted text and images are stored either way. Defaults to false: the original is kept so the user can download it and Squid can reprocess it. Without it, querySpreadsheetsWithAi cannot open a workbook context and a reupsert of the context fails, on the sync and the bulk lane alike.

    extractImages?: boolean

    Whether to extract images from the file, describe them with a vision model, and index those descriptions alongside the text. Defaults to TRUE — a document's charts, diagrams and scanned pages carry content the text layer does not, so a file context captures them unless asked not to.

    Note this is the opposite of AiKnowledgeBaseTextContextRequest.extractImages, which defaults to false: images referenced from a pasted markdown/HTML blob are usually layout chrome, and fetching them means following URLs the caller supplied.

    imageExtractionModel?: AiChatModelSelection

    The AI model to use for extraction, if specified.

    imageMinSizePixels?: number

    The minimum size for extracted images, if applicable.

    metadata?: AiContextMetadata

    Optional metadata associated with the context item.

    options?: AiKnowledgeBaseContextFileOptions

    General options for how to process the file.

    preferredExtractionMethod?: DocumentExtractionMethod

    The preferred method for extracting data from the document.

    scopes?: string[]

    Access scopes granting read access to this context, stamped onto every chunk it produces. Only meaningful on a knowledge base with accessControl.mode: 'enforced', where a caller retrieves the context only if they hold at least one of these scopes.

    Supplied here rather than through a follow-up call so a restricted context is never briefly readable by everyone between being indexed and being scoped. Omitted on a re-ingest keeps whatever scopes the context already had; an empty array makes it visible to every caller.

    Scopes are opaque strings the resolver and the ingest side agree on. The __ prefix is reserved.

    type: "file"

    Specifies the context type as 'file'.

    Settings
    Member Visibility

    On This Page
    Properties