This response starts a three-step upload flow. First, request this context from OpenSea. Second, call the returned method at the returned URL. For POST, add every fields entry unchanged as a multipart text field, then add a file part containing the bytes. The file part must be last. Let the HTTP library generate the multipart boundary; do not set the overall multipart Content-Type header yourself. POST storage uploads normally return 204. For PUT, upload the raw bytes, use only headers explicitly required by the endpoint, and expect 200. Treat any 2xx storage response as success. The URL and fields are short-lived sensitive credentials. Do not log, persist, alter, or put them in tickets. Third, after storage succeeds, pass the returned token to the documented OpenSea API endpoint. Do not use the token before the storage upload succeeds. Pass each token as media_token to POST /api/v2/drops/{slug}/items or POST /api/v2/drops/{slug}/prereveal-item, as media_tokens to POST /api/v2/drops/{slug}/items/media/save, or as media_token to PUT /api/v2/drops/{slug}/items/{token_id}. To save a drop's items in bulk, pass an upload_batch_id on the upload requests and save with POST /api/v2/drops/{slug}/items/media/save-batch, which takes filenames rather than tokens. Passing media_tokens to POST /api/v2/drops/{slug}/items/media/save is deprecated.