What to do when a post fails
The old behaviour was worse than the failure. A post scheduled for 17:55 would retry through the night and end up published at 04:30, which is a genuinely bad outcome for a post about your launch. Failures happen: a platform rejects a file, a token expires, an API has a bad afternoon. Forgefy retries briefly, then stops and tells you.
Step 1: Open the failed post. The card shows the real provider error.
Open the failed post. The card shows the real provider error.
Step 2: Fix the cause: reconnect the account, adjust the media, or shorten the caption.
Fix the cause: reconnect the account, adjust the media, or shorten the caption.
Step 3: Approve again. Forgefy re-books it with the provider.
Approve again. Forgefy re-books it with the provider.
Step 4: For a broken render specifically, use Fix visual, which reopens the editor rath…
For a broken render specifically, use Fix visual, which reopens the editor rather than generating a brand new AI image.
Good to know
Automatic technical retries are capped at roughly 30 minutes. After that it stops and waits for you. Your scheduled time is an immutable calendar reference. Retries never move it. Account-level errors, like an expired token, pause that whole platform instead of burning through every queued post. One probe per cycle checks whether it recovered. A post that has been stuck in flight for 45 minutes triggers a deep probe of the provider to find out what really happened. X rejects some video encodings. Forgefy attempts an automatic repair, including injecting a silent audio track when the file has…
Will a retry double-post?
No. Every publish takes an exclusive claim first, so a retry cannot run alongside the original attempt.
I deleted a post but it is still there.
Refresh once. Deletes are recorded locally so a stale realtime update cannot resurrect the card. If it survives a refresh, that is a bug.
Related