S3 APIs Integration

Orvanta’s s3 connection type works with AWS S3 as well as any S3-compatible object store — MinIO, Cloudflare R2, Backblaze B2, and similar services all speak the same API.

Add an S3 connection

The connection’s fields:

Property	Type	Description	Required
bucket	string	The bucket name	true
region	string	The bucket’s region	false
endPoint	string	The S3 endpoint (e.g. s3.amazonaws.com, or your MinIO/R2 host)	true
useSSL	boolean	Use TLS for the connection (defaults to true)	false
accessKey	string	The access key	true
secretKey	string	The secret key	true
pathStyle	boolean	Use path-style addressing instead of virtual-hosted-style (needed by most self-hosted S3-compatible stores)	false
port	integer	The endpoint port, if non-default	false

Once created, the connection can be passed as a parameter or fetched directly from scripts and flows like any other resource.

Workspace object storage

Workspace object storage is a Teams-and-above feature that lets scripts and flows read and write to S3, Azure Blob, or Google Cloud Storage without the runnable ever seeing the underlying credentials — Orvanta resolves the named storage resource and proxies the transfer, so access can be granted at the workspace level instead of handing out bucket secrets.

Instance object storage

An instance can additionally point worker dependency-cache storage at S3 (or an S3-compatible store), rather than re-downloading dependencies on every worker; the dependency cache is available on every tier. Job log archival to S3 is a separate Enterprise-tier capability that is implemented but held behind the beta toggle until it has been exercised against a live S3 deployment. This is configured separately from workspace-level storage and builds on the same connection.