AWS S3 n8n integrationAutomate AWS S3 with n8n.

A new bucket appears in the account behind the credential, and the workflow carries on with that name. Useful when each client needs its own storage space, not a shared prefix.

• Name: the name of the bucket to create, typed in or built from an expression, the n8n syntax that reads a value from the incoming item, such as {{ $json.client_slug }}.
• Region: in Additional Fields, the region the bucket is created in; left empty, n8n uses the region on the credential.
• ACL: the canned ACL applied to the bucket, from Private to publicRead.

One field is all this operation asks for, which makes it easy to wire at the end of a lifecycle: a bucket created for a single batch, removed once the batch is confirmed.

• Name: the name of the AWS S3 bucket to delete, usually carried from an earlier node with an expression like {{ $json.bucket }} rather than typed by hand.

Nothing is required here: n8n returns the buckets the credential can see, one item per bucket, ready for a loop or a filter downstream.

• Return All: switched on, n8n walks the pages of the API until every bucket is back.
• Limit: the maximum number of buckets returned while Return All stays off, which is plenty for a check inside the editor.

Inside a single bucket, this operation returns the keys it holds and can group them, which is the closest a workflow gets to browsing a folder tree.

• Bucket Name: the bucket to search in.
• Prefix: in Additional Fields, limits the response to keys starting with the string you give, a client folder for instance.
• Delimiter with Include Common Prefixes: the delimiter groups keys, and the option returns those grouped prefixes rather than the object contents.
• Start After: the key S3 starts listing from, to resume where a previous run stopped.

Copying writes the object to a second path and leaves the source alone. Both paths start with a slash, then the bucket, then the key, as in /bucket/my-image.jpg.

• Source Path: the source bucket and the key of the source object, separated by a slash.
• Destination Path: the destination bucket and the key to write, for example /bucket/my-second-image.jpg.
• Storage Class: in Additional Fields, where the copy lands, from standard to glacier or deepArchive.
• Metadata Directive: copy keeps the metadata of the source object, replace uses what the request provides.

Removing an object needs the bucket and the exact key, and nothing more unless the bucket keeps versions.

• Bucket Name: the bucket holding the object.
• File Key: the full key of the object, folder prefix included.
• Version ID: in Options, targets one version of the object instead of the current one.

The object comes back as binary data attached to the item rather than as text in a field, which is what the nodes downstream expect when they send or parse a file.

• Bucket Name: the bucket to read from.
• File Key: the key of the object to download.
• Put Output File in Field: the name of the output binary field the file is placed in, and the name every later node has to read.

Objects in a bucket come back one item at a time, and a folder key narrows the list before the workflow starts looping over it.

• Bucket Name: the bucket to read.
• Folder Key: in Options, restricts the listing to one folder instead of the whole bucket.
• Return All: leave it off while building, because the run then stops at Limit rather than walking a bucket with thousands of keys in it.

Whatever the previous node produced ends up in the bucket under the name you choose, as binary data or as plain text typed into the node.

• Bucket Name: the destination bucket.
• File Name: the key written in the bucket, hello.txt or a name built from the item.
• Binary File and Input Binary Field: turn Binary File on and name the input binary field holding the file; turn it off and File Content takes the text to store.
• Parent Folder Key: in Additional Fields, the folder the file is created in.

n8n writes a folder key into the bucket, so later operations can address files by folder instead of through one long flat key.

• Bucket Name: the bucket the folder belongs to.
• Folder Name: the name of the folder to create.
• Parent Folder Key: in Additional Fields, the folder you want to create it in, which is how a second level appears.
• Storage Class: also in Additional Fields, from standard to intelligentTiering or onezoneIA.

Two fields, no options, no room for a surprise: the bucket and the folder key, and the folder is gone.

• Bucket Name: the bucket that holds the folder.
• Folder Key: the key of the folder to remove, as it appears in the bucket.

Folders sitting in a bucket come back as items, which gives a workflow a map of the place before it decides where to write.

• Bucket Name: the bucket to inspect.
• Folder Key: in Options, the folder whose content gets listed instead of the bucket root.
• Fetch Owner: the owner field is absent from the response by default, so switch it on to get the owner with each key.