Skippr Docs

Appearance

How to Configure SFTP as a Source Connector ​

May 2026

SFTP source setup is easiest when the auth method and remote path are settled before the first automated download.

Short Answer ​

Configure SFTP as a source by setting source.kind to sftp, then providing host, username, and remote_path. Use either password or private_key_path for authentication, with port 22 as the default unless your server exposes SFTP on a different SSH port.

The clean first setup is usually private-key auth, one predictable directory or glob such as /data/*.json, and a user account that can list and read that location. If you must start with password auth, keep the secret in SFTP_PASSWORD so the connector definition stays portable across environments.

Why Teams Struggle with This ​

SFTP issues usually come from auth assumptions and path mismatches. The login may work interactively, but the automation user cannot read the folder you pointed Skippr at, or the remote glob does not match the filenames the export job actually produces.

  • The runner must reach the SFTP host on the configured SSH port.
  • The SSH user needs read access to the remote path and permission to list the directory if you use globs.
  • Private-key auth is usually a better long-lived setup than storing a password.
  • A vague remote_path can make troubleshooting harder than a precise directory or filename pattern.

How Skippr Handles It ​

Skippr keeps SFTP focused on the operator choices that matter most: host, port, auth method, and the exact remote path to read. That makes it practical for teams inheriting partner feeds or legacy export workflows where the file server already exists and the ingestion job simply needs to connect reliably.

The first working setup should be easy to confirm. Skippr can establish the SSH session, enumerate the path you asked for, and start downloading the expected files without extra wrappers or a separate file-transfer toolchain.

  • Support for password or private-key authentication.
  • Config that keeps remote_path explicit for repeatable file pulls.
  • Simple host and port controls aligned with standard SSH operations.
  • A good fit for partner data feeds and scheduled export drops.

What the First Useful Version Looks Like ​

Prefer private-key auth before the pipeline becomes business-critical, because rotating a key is usually cleaner than rotating shared passwords across environments.

For the first run, use a remote path that matches a small known set of files so you can validate ingestion without guessing whether the glob is too broad or too narrow.