A publish was refused
View as markdownWhat you see
Your assistant tells you the save was refused, and passes on why. The product says, for example:
is not among the uploaded files — nothing would be served at /
total publish exceeds the 5 GB limit for publishes with staged uploads
plan's limit, and it is reached —
— delete old artifacts or upgrade
and acts only there — a credential works in the workspace it was minted in
Why it happens
Every save is checked before anything is published, and a refused save changes nothing. An artifact needs an entry page — the file served at its address — so a save that names one it does not include, or leaves out the one it had, is refused. Files sent inside the conversation are capped at 25 MB each and 50 MB together; larger ones are staged first, up to 4 GB a file. Saves are counted per hour, and a plan has a number of artifacts and an amount of storage. A token works only in the workspace it was made in, so a script or a job saving into another workspace you belong to — your Personal, or a team's — is refused, and an artifact kept there is not found through it.
What to do
Ask your assistant to read the refusal back to you: it names the rule and the fix. For a missing entry page, ask it to include the page, or to name the new one. For large files, ask it to stage them (its guide is Large sites). When the hourly rate is reached, the refusal says when the next save is allowed. For the plan's limits, delete artifacts you no longer need, or upgrade. For a token made in another workspace, make one in the workspace the job saves into (that workspace → Settings → Credentials) and give the job that one; a job that saves into two workspaces needs a token from each.
Still stuck? Write to support with what your assistant was asked and what it said.