Uploading & Downloading

The CLI wraps the presign → PUT → confirm flow, so object bytes move directly between your machine and object storage — the API only issues signed URLs and records the result.

Upload

medialocker media upload ./hero.mp4 --bucket marketing

--bucket accepts a bucket name or id. Useful options:

Option	Effect
--bucket <id|name>	Destination bucket (required)
--key <key>	Object key; defaults to the filename (single file only)
--content-type <type>	Override the detected content type
--tag <tag>	Apply a tag; repeat for several

Upload several files at once:

medialocker media upload ./shots/*.jpg --bucket product --tag catalog

Behind the scenes the CLI requests a presigned URL, PUTs the bytes, and calls confirm to record the true size and enqueue derivatives. Files ≥ 100 MB automatically switch to the multipart flow — parts are uploaded and completed for you.

With --json, a successful upload prints the confirmed object:

{
  "objectId": "a17c…",
  "key": "hero.mp4",
  "size": "734003200",
  "bucketId": "b3f1…",
  "status": "confirmed"
}

Download

medialocker media download <objectId> --out ./hero.mp4

Option	Effect
--out <path>	Output path; defaults to the object key's basename
--expires <seconds>	Presigned URL lifetime (60–604800)

Presigned URLs

To hand a URL to another tool instead of downloading, print one:

medialocker media url <objectId> --expires 3600     # short-lived download URL
medialocker media thumbnail <objectId>              # thumbnail/poster URL

Scriptable round-trips

Both commands emit just the URL on stdout, so they pipe cleanly:

curl -o out.jpg "$(medialocker media url <objectId>)"

See also

• Presigned Uploads — the underlying single-shot and multipart flow
• Command Reference — the rest of the media commands