If you're looking for the "approval" or "review" step in an auto-posting workflow before you trust it, it's worth saying plainly: there isn't a dedicated one. That's not a gap to route around — it's a deliberate design, and understanding the actual mechanism matters more than searching for a button that doesn't exist.
What actually functions as review
The Trigger node has two modes: manual and timer. On manual, running the workflow drafts the post and stops there for you to look at — the slides, the caption, whatever the Apply Template step produced — before anything reaches the Post node. That look is the review. There's no separate gate node sitting between Apply Template and Post; the review happens by choosing when you press run, not by a checkpoint the workflow enforces on its own.
Why this is worth knowing before you build anything
If you assume there's an automated safety net catching bad output, you'll switch to a timer too early, on the assumption something else is watching for mistakes. There isn't. The actual safety net is however many manual runs you do first — see debugging a failed automation run for what a bad run actually looks like when you catch it this way, and the pre-launch checklist for what to verify before flipping to a timer at all.
How many manual runs is "enough"
There's no fixed number, but the honest signal is boredom, not a count — once you're running it manually and consistently thinking "yeah, that's right" without needing to correct anything, you've probably seen enough variation in the underlying data to trust it unattended. A data source that only shows one kind of value in your first three test runs hasn't actually been stress-tested yet, even if all three looked fine.
Higher-stakes content deserves a longer manual phase
Not every format carries the same risk if it's wrong — financial or factual content is worth a longer manual run-in than a low-stakes daily special, because the cost of a mistake is genuinely different. Match how long you stay on manual to what's actually at risk if this specific automation gets something wrong, not to a one-size-fits-all rule.
After you switch to a timer
Moving to a timer doesn't mean review is over forever — it means the per-run check moves from before publish to after. Spot-check the account periodically, and if you edit the workflow later, drop it back to manual for a few runs before trusting the edited version unattended again, the same caution you'd apply to a brand-new automation.