Paperless MCP

Get one correspondent by id.

Get one custom field definition by id, with its select options.

Get one document's metadata, notes and custom field values, without its text. Use get_document_content for the text.

Read a document's text, up to 20,000 characters per call. A longer document comes back in sections: each opens with a marker naming the range returned, the full length and the offset of the next section. Text that fits is returned whole, with no marker. Args: max_chars: Most characters to return in this call. offset: Character position to start from; pass the offset a marker names to read the next section.

List the changes made to a document: who changed which field, and when.

Get a document's file details: original and archived file names, sizes, checksums and MIME type.

List the notes on a document, with their full text.

Get Paperless's suggested tags, correspondent, document type and dates for a document.

Get a small image of a document's first page.

Get one document type by id.

Check whether a newer Paperless-NGX release exists; returns the newest release. That is the newest release published upstream, not the version installed on this instance, with a flag saying whether it is newer. Use get_server_info for the installed version.

Get one saved view by id, with its filter rules and sort order.

Get one share link by id, with its document, expiry and URL.

Get counts for the whole archive: documents, inbox documents, tags, correspondents, types, characters and file types.

Get one storage path by id, with its path template.

Get one tag by id.

Get one background task by id, with its status and result; returns null if there is none. Args: task_uuid: Task id, as returned by upload_document or listed by list_tasks.

List correspondents; returns one page with their ids and names. Each carries last_correspondence, the date of its newest document.

List custom field definitions; returns one page with their ids, names and types.

List document types; returns one page with their ids and names.

List documents filtered by their metadata; returns one page of metadata. Use search_documents to find documents by the words in them. Results leave out each document's text, note text and custom field values; get_document_content and get_document return them. Args: tags: Keep documents that carry any of these tag ids. correspondent: Keep documents from this correspondent id. document_type: Keep documents of this document type id. storage_path: Keep documents in this storage path id. custom_field: Keep documents that have this custom field.

List saved views, the named document filters users keep in Paperless; returns one page.

List public share links for documents; returns one page with each link's document, expiry and URL. Args: document_id: Keep only the links to this document.

List storage paths, the folder layouts Paperless files documents into; returns one page.

List tags; returns one page with their ids and names.

List Paperless background tasks such as consuming uploads and reindexing; returns one page, newest first. Args: status: Keep tasks in this state. task_type: Keep tasks of this kind; bulk_update is the reindex that bulk_edit_documents' metadata operations start. acknowledged: Keep only acknowledged (true) or unacknowledged (false) tasks. include_acknowledged: Also list tasks already acknowledged; by default only unacknowledged ones are listed.

Search the full text and metadata of documents; returns one page, best match first. Use list_documents to filter by tag, correspondent or type alone. Results leave out each document's text, note text and custom field values; get_document_content and get_document return them. Args: query: Words that must all appear, in any order, in a document's text, title, correspondent, type or tags. Also accepts AND and OR, field terms such as tag:unpaid or created:[2024 to 2025], and notes.note:word for note text. Pass "" with more_like. more_like: Document id; return docu…

Change a correspondent's name or matching rule; returns the updated correspondent. Args: patch: Only the fields to change.

Rename a custom field or change its options; returns the updated field. Args: patch: Only the fields to change.

Change a document's metadata; returns the updated document without its text. Use bulk_edit_documents to change many documents at once. Args: patch: Only the fields to change.

Change a document type's name or matching rule; returns the updated document type. Args: patch: Only the fields to change.

Change a tag's name, colour, inbox flag or matching rule; returns the tag. Args: patch: Only the fields to change.

Upload a file for Paperless to consume; returns the consume task's id. The document appears once the task finishes; wait_for_task waits for it and its result names the new document. For large files, use create_upload_link where it is available. Args: filename: File name, with the extension that tells Paperless its type. content_base64: The file's bytes, base64-encoded. title: Title; omit to let Paperless derive one. correspondent: Correspondent id. document_type: Document type id. tags: Tag ids to add. created: Creation date, YYYY-MM-DD or ISO 8601. archive_se…

Wait for a background task to finish; returns the task with its final status and result. Args: task_uuid: Task id, as returned by upload_document or listed by list_tasks. timeout_seconds: Longest time to wait, up to 600 seconds.