Overview

urlInput fetches one or more remote files over HTTP or HTTPS and converts them into workflow file objects. It is useful when the workflow should ingest public files without a manual upload step.

Configuration / Settings

Setting	Type	Required	Details
URLs	string[]	Yes	The executor reads `config.urls`. Empty values are ignored.
Variables in URLs	string template	No	URL strings can include `{{node_id.field}}` expressions before fetch.

Inputs Expected

No upstream input is required.
Each URL must resolve to http: or https:.
Private-network, loopback, metadata-service, and internal hostnames are blocked by the executor.

Variables Exported / Outputs

Output shape: Blob[]

Metadata recorded in execution logs:

outputFileIds
fetchedUrls
errors when some URLs fail but at least one succeeds

Common Use Cases

Pull a PDF from a signed URL, then compress or split it.
Download generated files from an external system before emailing or archiving them.
Combine URL-based files with uploaded files in a later merge step.

Edge Cases

The node throws if every URL fails.
If some URLs fail and at least one succeeds, the node still completes and records the failures in metadata.
Filenames are inferred from Content-Disposition first, then from the URL pathname.
Only direct file downloads work well. HTML landing pages are fetched as text blobs unless the remote server returns the real file.