Only one automation replies

When several automations could answer, exactly one does. How the winner is chosen.

If two or more of your automations match the same comment or message, only one of them answers. The others do nothing and are recorded as skipped.

This is deliberate. Without it, one comment could receive two public replies from two automations, under your own post, where everyone can see it.

How the winner is chosen

Three tests, in order. The first one that separates them decides.

  1. Specific posts beat all posts. An automation that names the post wins over one watching everything.
  2. Exact beats contains, and contains beats any. The narrower match mode wins.
  3. If they are still tied, the older automation wins.

What this means for how you build

A catch-all is safe

An automation set to all posts with match mode Any answers everything that nothing more specific claimed. That is almost always what you want a catch-all to do: it is the fallback, not the winner.

Narrow beats broad, so build the narrow one

If you want a particular post to get a special reply, create an automation for that post. It will win over your general one automatically. You do not need to exclude the post from the general automation.

Two similar automations will surprise you

If two automations both watch all posts with match mode Contains, and a comment contains both their keywords, one of them stays silent. Different keywords do not make automations independent, because one comment can contain both.

The warning when you publish

When you publish an automation that watches a post another live automation already watches, dmrabbit tells you. It is a warning, not a refusal: overlapping automations are sometimes exactly what you want.

The warning is about posts, not keywords. It cannot tell you whether your keywords will collide, because that depends on what people actually write.

Finding out which one answered

Every automation records what it did. An automation showing skipped matched, but another one answered instead.