Prompt-To-Campaign Builder

Build campaigns from a prompt.

Prompt the full campaign. Then prompt each message.

Tell Tuco the audience, trigger, and outcome you want. It drafts the step tree, writes AI or manual messages, treats each AI message as its own prompt when you want tighter control, pulls live HubSpot and GoHighLevel context, and tests the exact path on a real lead before launch.

Prompt the campaignPrompt each messageReal-lead simulatorCRM stop checks

Last updated:

Prompt in

Describe the motion.

“Build a 3-step follow-up for GHL no-shows. Stop when HubSpot says booked.”

Tree out

Steps write themselves.

Manual and AI steps can mix in one path. Each AI step can carry its own prompt. Branches and fallbacks are included.

Live data

Real CRM checks.

The builder can look at HubSpot notes, GHL activity, tags, properties, and custom API responses.

Before launch

Simulate the path.

Pick a lead. Fast-forward waits. Watch exactly what sends, what branches, and what stops.

Campaign draft

Prompt, generate, simulate, refine.

Live
System

Lead source: GoHighLevel no-show list · HubSpot stage: SQL

Step 1: AI writes a short follow-up using the lead name, company, and last meeting context.

AI step

System

Branch after 48 hours if no reply. Stop if a HubSpot note contains “booked” or GHL tag = won.

Fallback SMS and fallback copy are attached to every AI step before launch.

Safety layer

1 panel
Prompt and preview

Write with AI on the left. Inspect the live campaign structure on the right.

Per step
AI where it matters

Use AI for the opener, the follow-up, or every branch. Manual and AI can coexist.

Live
HubSpot and GHL checks

The simulator can evaluate real notes, tags, activities, and properties before you activate.

Built in
Fallback protection

Every AI step can carry manual fallback copy and optional SMS fallback if iMessage fails.

Campaigns. Without busywork.

Tell it the play.
Prompt to structure

The builder asks clarifying questions, pulls workspace context, and turns a plain-English idea into a real step tree.

Each message is a prompt.
🧠
AI per step

Switch any message to AI mode and guide it step by step. Each message can personalize itself from lead data, conversation history, tags, notes, and CRM context.

Test the actual path.
🧪
Live simulation

Choose a real lead, reply as that lead, and watch branches, waits, stop conditions, and fallback logic fire in order.

Stop when reality changes.
🛑
Dynamic stop logic

Messages can stop on HubSpot notes, GHL tags, CRM property changes, Tuco replies, or external API results.

See CRM signals
Give the bot a tool.
📡
Custom tools

Wire in external APIs so the campaign can check inventory, credit, lifecycle, or anything else before it sends. The same custom tool can also be reused by AI SDRs.

See custom tools
Fallbacks stay attached.
🛟
Fail safely

If the model times out, costs spike, or a provider flakes, fallback copy keeps the step useful instead of blank.

Workflow

How it flows.

Describe the campaign. Watch it get sharper with every pass.

01

Describe the campaign.

Name the audience, the trigger, and the outcome you want. Tuco starts with the shape, not a blank canvas.

02

Pull in live context.

The builder can look up tags, notes, properties, and workspace tools before it writes the next branch.

03

Simulate the path.

Select a lead. Skip time. Trigger replies. Confirm where the path sends, waits, or stops.

04

Activate with guardrails.

Fallback copy, SMS backup, and stop conditions stay attached when the campaign goes live.

Build the campaign. Before lunch.

Your team should spend time tuning the motion, not dragging branches for an hour. Tuco gives you the structure, the simulation, and the stop logic in one place.

Get StartedMini self-serve, Growth demos