client.kb_documents

Knowledge-base documents assistants answer from. These methods are on client.kb_documents, where client is a MoreVoice client (see the Python SDK). On AsyncMoreVoice the same methods are awaited. Each one returns the response object and raises an exception when the API answers with an error.

`create()`

Add a knowledge-base document. Adds a document from a file (JSON with base64, or multipart/form-data with a file part, ≤ 25 MB), a public URL, or text, and indexes it in the background: answers 202 with the document pending. Poll it until status is ready (searchable) or error. Beyond the plan’s knowledge-base storage: 402 plan_limit.

def create(self, body: _m.KbDocumentCreateParamsInput | _m.KbDocumentsCreateFilesBody | Mapping[str, Any] | Unset = UNSET, *, more_voice_version: str | Unset = UNSET, idempotency_key: str | Unset = UNSET) -> _m.KbDocument

client.kb_documents.create() · await async_client.kb_documents.create() · POST /kb/documents · API reference

Request body

Field	Type	Required	Description
`file`	`object`	no	A file to upload. Or send the request as multipart/form-data with a `file` part.
`format`	`"md" \| "txt" \| "csv"`	no	With `text`: how to read it (default md).
`text`	`string`	no	Text to index as it is.
`title`	`string`	no	Default: the file’s name, the page’s title, or (text) “Document”.
`url`	`string`	no	A public web page or file (http/https) to fetch and index. It is fetched again on reingest.

Returns

KbDocument:

Field	Type	Description
`id`	`string`	The document’s ID.
`object`	`"kb_document"`	Always `kb_document`.
`livemode`	`boolean`	`true` in live mode, `false` in test mode.
`title`	`string`	—
`kind`	`"pdf" \| "docx" \| "md" \| "txt" \| "csv" \| "xlsx" \| "html" \| "url"`	The format it was read as; `url` for a web page.
`source`	`string`	The uploaded file’s name, or the URL.
`status`	`"pending" \| "processing" \| "ready" \| "error"`	pending → processing → ready (searchable) \| error. Poll the document, or follow `kb.document.*` events.
`error`	`string \| null`	Why ingestion failed (status error).
`bytes`	`integer`	—
`chunk_count`	`integer`	Searchable passages the document was split into.
`original_stored`	`boolean`	The original file is kept (encrypted): GET …/content downloads it and reingest re-reads it.
`created`	`string`	An ISO-8601 timestamp in UTC.
`updated`	`string`	An ISO-8601 timestamp in UTC.

Example

from morevoice import MoreVoice

client = MoreVoice()  # MOREVOICE_API_KEY from the environment

kb_document = client.kb_documents.create({
    "text": "# Opening hours\nSunday–Thursday 08:00–18:00, Friday 08:00–13:00.",
    "title": "Opening hours",
})
print(kb_document)

`delete()`

Delete a knowledge-base document. Deletes the document, its passages and its stored original. Assistants and the copilot stop answering from it at once.

def delete(self, id: str, *, more_voice_version: str | Unset = UNSET, idempotency_key: str | Unset = UNSET) -> _m.DeletedKbDocument

client.kb_documents.delete() · await async_client.kb_documents.delete() · DELETE /kb/documents/{id} · API reference

Parameters

Name	Type	Required	Description
`id`	`str`	yes	A kb document ID (`kbd_…`).

Returns

DeletedKbDocument:

Field	Type	Description
`id`	`string`	A kb document ID (prefix `kbd_`).
`object`	`"kb_document"`	Always `kb_document`.
`deleted`	`true`	—

Example

from morevoice import MoreVoice

client = MoreVoice()  # MOREVOICE_API_KEY from the environment

deleted_kb_document = client.kb_documents.delete("kbd_7Hk2Lm9Qp")
print(deleted_kb_document)

`list()`

List knowledge-base documents. Returns a page of KbDocument objects, newest first. Pass next_cursor as starting_after for the next page; the SDKs iterate every page for you.

def list(self, *, limit: int | Unset = 20, starting_after: str | Unset = UNSET, ending_before: str | Unset = UNSET, status: _m.KbDocumentsListStatus | Unset = UNSET, more_voice_version: str | Unset = UNSET) -> AsyncPage[_m.KbDocument]

client.kb_documents.list() · await async_client.kb_documents.list() · GET /kb/documents · API reference

Parameters

Name	Type	Required	Description
`limit`	`integer`	no	How many objects to return, 1–100 (default 20).
`starting_after`	`string`	no	A cursor (`next_cursor`) or object ID: return the objects after it (older).
`ending_before`	`string`	no	A cursor or object ID: return the objects before it (newer).
`status`	`"pending" \| "processing" \| "ready" \| "error"`	no	Only documents in this status.

Returns

A page of results (KbDocumentList): data, has_more and next_cursor.

Example

from morevoice import MoreVoice

client = MoreVoice()  # MOREVOICE_API_KEY from the environment

for kb_document in client.kb_documents.list():
    print(kb_document)

`reingest()`

Re-index a knowledge-base document. Reads the document again — a URL is fetched again, a file is re-read from its stored original — and re-indexes it in the background (202). 409 when the original of an uploaded file is not stored: upload it again.

def reingest(self, id: str, *, more_voice_version: str | Unset = UNSET, idempotency_key: str | Unset = UNSET) -> _m.KbDocument

client.kb_documents.reingest() · await async_client.kb_documents.reingest() · POST /kb/documents/{id}/reingest · API reference

Parameters

Name	Type	Required	Description
`id`	`str`	yes	A kb document ID (`kbd_…`).

Returns

KbDocument:

Field	Type	Description
`id`	`string`	The document’s ID.
`object`	`"kb_document"`	Always `kb_document`.
`livemode`	`boolean`	`true` in live mode, `false` in test mode.
`title`	`string`	—
`kind`	`"pdf" \| "docx" \| "md" \| "txt" \| "csv" \| "xlsx" \| "html" \| "url"`	The format it was read as; `url` for a web page.
`source`	`string`	The uploaded file’s name, or the URL.
`status`	`"pending" \| "processing" \| "ready" \| "error"`	pending → processing → ready (searchable) \| error. Poll the document, or follow `kb.document.*` events.
`error`	`string \| null`	Why ingestion failed (status error).
`bytes`	`integer`	—
`chunk_count`	`integer`	Searchable passages the document was split into.
`original_stored`	`boolean`	The original file is kept (encrypted): GET …/content downloads it and reingest re-reads it.
`created`	`string`	An ISO-8601 timestamp in UTC.
`updated`	`string`	An ISO-8601 timestamp in UTC.

Example

from morevoice import MoreVoice

client = MoreVoice()  # MOREVOICE_API_KEY from the environment

kb_document = client.kb_documents.reingest("kbd_7Hk2Lm9Qp")
print(kb_document)

`retrieve()`

Retrieve a knowledge-base document. Returns the KbDocument object. Answers 404 with the code resource_missing when nothing has this ID in this organisation and mode.

def retrieve(self, id: str, *, more_voice_version: str | Unset = UNSET) -> _m.KbDocument

client.kb_documents.retrieve() · await async_client.kb_documents.retrieve() · GET /kb/documents/{id} · API reference

Parameters

Name	Type	Required	Description
`id`	`str`	yes	A kb document ID (`kbd_…`).

Returns

KbDocument:

Field	Type	Description
`id`	`string`	The document’s ID.
`object`	`"kb_document"`	Always `kb_document`.
`livemode`	`boolean`	`true` in live mode, `false` in test mode.
`title`	`string`	—
`kind`	`"pdf" \| "docx" \| "md" \| "txt" \| "csv" \| "xlsx" \| "html" \| "url"`	The format it was read as; `url` for a web page.
`source`	`string`	The uploaded file’s name, or the URL.
`status`	`"pending" \| "processing" \| "ready" \| "error"`	pending → processing → ready (searchable) \| error. Poll the document, or follow `kb.document.*` events.
`error`	`string \| null`	Why ingestion failed (status error).
`bytes`	`integer`	—
`chunk_count`	`integer`	Searchable passages the document was split into.
`original_stored`	`boolean`	The original file is kept (encrypted): GET …/content downloads it and reingest re-reads it.
`created`	`string`	An ISO-8601 timestamp in UTC.
`updated`	`string`	An ISO-8601 timestamp in UTC.

Example

from morevoice import MoreVoice

client = MoreVoice()  # MOREVOICE_API_KEY from the environment

kb_document = client.kb_documents.retrieve("kbd_7Hk2Lm9Qp")
print(kb_document)

`retrieve_content()`

Download a document’s original file. The uploaded original, as it was uploaded (its own content type). 404 for URL documents and when the original is not stored (original_stored: false). Every download is audited.

def retrieve_content(self, id: str, *, more_voice_version: str | Unset = UNSET) -> File

client.kb_documents.retrieve_content() · await async_client.kb_documents.retrieve_content() · GET /kb/documents/{id}/content · API reference

Parameters

Name	Type	Required	Description
`id`	`str`	yes	A kb document ID (`kbd_…`).

Returns

Nothing, on success.

Example

from morevoice import MoreVoice

client = MoreVoice()  # MOREVOICE_API_KEY from the environment

data = client.kb_documents.retrieve_content("kbd_7Hk2Lm9Qp")
print(data)