Generate an AI Dashboard as a Run User
POST/run/:runURL/dashboard/:dashboardID/generate
Build or change a signed-in TagoRUN user's own copy of an AI Dashboard from a prompt.
This is the run-user prompt bar route. The user's grants decide what it can do, and it never acts on another user's copy.
Experimental. This feature may change or be removed.
Send the TagoRUN user's own token in the Authorization header. The Anonymous User, profile tokens,
and other token types are refused. The user needs:
- Dashboard → Access and Dashboard → AI dashboard generation on the dashboard
- AI → Access
- AI Chat enabled for the portal (TagoRUN → AI), and AI enabled for the profile
AI → Read device data and Device → AI access are not needed here: they only let the TagoRUN AI Chat look up devices. The generator finds the user's devices through Device → Dashboard access.
A user without a copy starts from the dashboard itself when per_user is off, or from an empty
dashboard when it is on. The first change saves their copy. See
Per Run User Generation.
Each prompt counts once against the profile's TagoRUN AI requests on the free model; paid models follow the profile owner's billing mode. See TagoAI Service.
The prompt and history limits are those of Generate AI Dashboard. Only
one generation runs at a time on each copy. Refusals keep the status, message, and Retry-After header
the generator answered with.
Request
Responses
- 200
- 400
- 404
- 429
- 503
- 504
The generation finished. changed is false when the result matched the stored copy: nothing was saved, and changes is empty.
Invalid body, a token that is not a signed-in TagoRUN user's, a run URL of another application, AI Chat off for the portal or AI off for the profile, or a user without AI → Access or Dashboard → AI dashboard generation on the dashboard ("Authorization denied").
The dashboard is not an AI Dashboard of this profile, or the user has no Dashboard → Access to it.
The user's copy is already generating, the profile already runs three generations, or the TagoRUN AI requests allowance is used up. The Retry-After header, when present, gives the seconds to wait.
The generator could not be reached, or did not answer in time.
The generation did not finish within 120 seconds.