Tools and runs
Run-it-now buttons for the work that normally happens on a schedule. Studio, then Tools.
The actions
| Action | What it does |
|---|---|
| AI planner run | Drafts a batch of posts, a reel and a carousel as backlog drafts |
| Comments scan | Scans your sources and drafts comments |
| Comments deep scan | The same at maximum depth. For a newly added source |
| Social token refresh | Refreshes connection tokens now |
| Research sync | Syncs tracked channels now |
The AI planner and both comment scans need a content profile. Without one they refuse with "No profile configured for this team yet".
Runs
Each click creates a run with a live console you can watch.
| Status | Meaning |
|---|---|
queued | Waiting for a worker |
running | In progress |
finished | Done |
failed | Check the console output |
AI runs take minutes rather than seconds. That is normal.
Run history pages as you scroll, so you can look back at what a previous run did.
What is not here
Carousel rendering and repurposing live on the draft, since they act on one specific draft.
Monthly reports run on the 1st automatically, and you read the output under Insights.
When to use it
AI planner run when your backlog is empty and you want a week drafted to react to. Everything it produces is a draft you edit.
Comments deep scan exactly once, after adding a subreddit, to backfill. Then leave it to the schedule.
Social token refresh when a connection shows needs_reauth and you want to check whether it recovers before reconnecting.
Research sync when you have just tracked a channel and do not want to wait 6 hours.
Everything is a web action
There is no CLI you are expected to run to operate Postlyra. The artisan commands behind these buttons exist for the scheduler, and the buttons dispatch the same jobs.
The one exception is the postlyra CLI, which is a different thing: it runs on your machine to mine your local git history, precisely because the server cannot read it.