capabilities.work

Upsert a Document and Search the Data Source

A 3-step recipe with Dust. We can help you get access.

Recipe · 3 steps

Upsert a Document and Search the Data Source

with Dust

Upsert a document, wait for indexing, and search the data source.

  1. Upsert the document into the data source synchronously so it is queued for indexing.
  2. Check how many upsert workflows are still running for this data source. Continue polling while any are running, otherwise proceed to search.
  3. Run a semantic search against the data source to confirm the upserted document is retrievable.
What you'll need to have ready
  • apiToken: Dust API key used as the Bearer token.
  • wId: The workspace identifier.
  • spaceId: The space identifier containing the data source.
  • dsId: The data source identifier.
  • documentId: The document identifier to upsert.
  • title: Title of the document.
  • text: Plain text content of the document.
  • query: The search query to run after indexing.
  • topK: Number of search results to return (defaults to 5).
For your developer

Arazzo workflow dust-tt:upsert-document-and-search, 3 operations: {$sourceDescriptions.datasourcesApi.url}#/paths/~1api~1v1~1w~1{wId}~1spaces~1{spaceId}~1data_sources~1{dsId}~1documents~1{documentId}/post → {$sourceDescriptions.datasourcesApi.url}#/paths/~1api~1v1~1w~1{wId}~1spaces~1{spaceId}~1data_sources~1{dsId}~1check_upsert_queue/get → {$sourceDescriptions.datasourcesApi.url}#/paths/~1api~1v1~1w~1{wId}~1spaces~1{spaceId}~1data_sources~1{dsId}~1search/get. Specs on APIs.io.

More recipes with Dust

Approve a Pending Agent Tool ActionAttach a File and Ask an AgentConverse with an AgentCreate a Table and Upsert RowsDiscover Spaces and Data SourcesFind Agent and Start a ConversationImport an Agent and Start a ConversationMessage an Agent and Submit FeedbackPost Follow-up Message and PollRun an App and Poll for Completion