Deploy rules: when the agent runs

The Deploy tab defines when this agent runs on an inbound message.
Think of it as the door policy — who gets let in.

The mental model

• Every workspace can have multiple agents
• When an inbound message arrives, we evaluate each agent's deploy rules
• The first agent whose rules match catches the message
• If no agent matches, the message goes unanswered (intentional — better than

Rule shape

Each rule is one or more groups. The boolean logic is:

• Within a group → AND. Every condition in the group must match.
• Between groups → OR. Any one group matching is enough.
• Each condition → NOT (optional). Click the NOT toggle on a condition
• Within a condition → OR. List multiple values and any one matches.

Example — one group, two conditions:

Contact has tag in [hot-lead, vip]
AND
Contact does NOT have tag [bot, do-not-contact]

Reads as: "Run this agent for any contact tagged hot-lead or vip, as
long as they don't also have bot or do-not-contact."

Example — two groups (OR):

Group 1:
  Contact has tag [enterprise]
  AND Contact in pipeline stage [closing]
OR
Group 2:
  Contact has tag [high-intent]
  AND NOT Contact has tag [cold]

Reads as: "Run this agent for enterprise deals in closing stage — OR —
high-intent contacts who aren't tagged cold."

Condition types

• All inbound messages — catch-all, matches everything (doesn't
• Contact has tag — picks from your GHL tags (multi-select, supports NOT)
• Contact in pipeline stage — pipeline stage ID (multi-value, supports NOT)
• Message contains keyword — keyword match against the inbound

Priority order

Rules are evaluated lowest-priority number first. By convention:
- Very specific rules → priority 10–50
- Catch-all fallback → priority 999

This way your "hot-lead" agent catches its contacts before the generic
agent scoops them up. A rule that's purely ALL inbound messages lands
at priority 999 automatically; anything more specific (including ALL
combined with a tag filter) lands at priority 10 so it can outrank a
plain catch-all.

Design tips

• Keep rules layered. One specific rule per segment, one catch-all.
• Avoid overlap. If two agents could both match, the first wins by
• Use NOT to carve out exceptions. Instead of writing three separate
• Use OR groups for "either of these makes sense". A nurture agent
• Between whole rules is still OR. Rule 1 fires OR rule 2 fires — not