site AND kind html (a single HTML page has a one-entry manifest); any other kind stores one opaque object with no addressable paths and is refused with 400. Isolation is IDENTICAL to GET /sites/{pageId}: see PageNotFound below. Refused with 403 if the subject is not in good standing (see Forbidden) — this applies to reads as well as writes.curl --location 'https://upload.tf/api/v1/sites/a1b2c3d4e5/files' \
--header 'x-api-key: <api-key>'{
"success": true,
"data": {
"files": [
{
"path": "string",
"sizeBytes": 0,
"contentType": "string",
"isEntry": true
}
],
"fileCount": 0,
"totalBytes": 0
}
}