Skip to main content

​
A provider is missing from Add connection

Check provider setup. All required values must be present. Partial configuration fails startup. Confirm callbacks match the public API origin and that your provider application has the requested approval.

​
A video stays in processing

Check the worker’s health and logs. Video preparation needs FFmpeg and writable temporary space. The worker image supplies FFmpeg and Compose reserves 256 MiB for /tmp. A failed normalization returns a failed media status; re-upload a valid MP4/MOV rather than forcing a row to ready.

​
Uploads fail in the browser

Check the exact UI origin in bucket CORS, the private bucket name, region, endpoint and bucket-scoped credential. Storage documents safe prefixes. An API health check does not exercise CORS or signed uploads.

​
A schedule does not advance

Check the UTC schedule time, account time-zone preferences and worker readiness. A healthy UI does not imply a running worker. Look at delivery details rather than repeatedly submitting the same post.

​
Delivery is Partial or Uncertain

Open delivery details. Confirmed successes are kept; inspect the failed account or provider page. An uncertain response may already correspond to a public post. Verify the provider before intentionally making another posting request. Check both public origins, HTTPS, session cookies and the current role. Role changes revoke human-owned MCP grants; reconnect through OAuth and select the access level again. Do not add a static MCP bearer token as a workaround.

​
First setup asks for a secret

Read the API’s initial startup log in your private operator session. The setup secret proves installation access. Existing teams cannot be initialized again. Use the operator recovery procedure for a lost administrator account.