Lifecycle Email Agent
Writes the sequence, and tells you which segment you may not email
- 8hard rules
- Things it must never do, enforced in every run
- 5handover rules
- Conditions where it stops and gives the work to a person
- 8output fields
- A structured result, not a paragraph you have to re-read
- ≤5steps per run
- Bounded work: at most 4 tool calls, enforced by the runtime
Read-only by design. This agent has no path to sending, spending, publishing or committing anything. Where that limit is the product, removing it would remove the reason to trust the output.
Overview
"Personalise emails with AI" is one of the most-copied automations there is, and every version treats a list as a resource rather than as people who each did or did not give permission.
This drafts the sequence — onboarding, re-engagement, renewal, a campaign — from what you actually know about the segment, in your own voice, with each message doing one job rather than three.
And before the copy, it reports what the copy cannot fix: segments with no recorded lawful basis, addresses suppressed elsewhere that this sequence would reach again, people already inside another live sequence who are about to receive two things at once, and messages missing the identification and unsubscribe elements that marketing mail requires. Sending without a basis is enforced in the EU and UK, one complaint run can cost a domain its reputation, and no amount of good copy survives either.
Benefits
Nothing sent you should not have sent
The failure that costs a sending domain, not a campaign.
Sequences written in an afternoon
The work that is always planned and rarely done.
Fewer unsubscribes
Mostly by not mailing people twice in a day.
What it does
Checks the basis before the copy
A segment with no recorded lawful basis is reported first. No sequence fixes that.
Catches cross-list suppression
Someone who unsubscribed from one list and is about to be reached through another — the exact failure regulators write about.
Finds collision with live sequences
Two automations reaching the same person in one day is how a list gets trained to ignore you.
One job per message
A message with three calls to action has none, and it is the commonest fault in a drafted sequence.
Writes from what you know
Personalisation from your data, never from an invented assumption about the recipient.
Sends nothing
Drafts and findings. Queuing a sequence is a decision with a domain's reputation attached.
What this changes
What this changes about your week
- It takes over
- Writing lifecycle sequences and checking who may actually be emailed.
- Done today by
- A marketer, and in practice nobody checks the second part.
- How often
- Every campaign and every lifecycle stage.
- Typical manual effort
- A multi-step sequence written properly is days.
Why Marketing in particular. Sending without a lawful basis is enforced in the EU and UK, and one complaint run can cost a sending domain its reputation. No amount of good copy survives that.
Runs unattended
Started by you or by an event, and it finishes on its own. Nothing waits for someone to be at a desk.
The same standard every time
The two-hundredth item is held to the bar the first one was. Consistency is the part people cannot sustain.
It cannot act on its own
Lifecycle Email Agent has no path to sending, spending or committing. That limit is why its output is safe to act on.
How it runs
This agent runs server-side through the PROMIVO runtime. Each run is logged step by step and every tool call is permission-checked before it executes.
- 1InputValidated against the agent input schema
- 2Modelclaude-sonnet-5 with automatic fallback
- 3Toolssearch_knowledge, extract_structured_data
- 4PermissionsRuns READ actions automatically; anything higher waits for approval
- 5OutputStructured result matching the output schema
- 6LoggingTokens, duration and every step recorded
Demo
Demo dataIllustrative sample output, abridged.
Input
{
"goal": "Re-engage lapsed trials",
"segment": {
"name": "Trial, no login 60d",
"size": 4120
}
}Output
{
"escalate": true,
"sequence": [
{
"day": 0,
"body": "…",
"oneJob": "Get them back to the one thing they started and abandoned.",
"subject": "The export you set up never finished",
"missingElements": [],
"personalisationUsed": [
"last incomplete action"
]
}
],
"basisCheck": {
"detail": "1,340 of 4,120 records carry no consent timestamp and no record of a customer relationship. Nothing was drafted for those.",
"status": "unclear"
},
"disclaimer": "Drafts and findings from your own records. Not legal advice on marketing rules, and nothing has been sent or queued.",
"totalFrequency": "3 messages over 11 days.",
"escalationReason": "A third of the segment has no recorded lawful basis, and 890 people would receive two sequences on the same day.",
"sequenceCollisions": [
"890 of this segment are also in the live 'Product update' sequence and would receive two emails on day 3."
],
"suppressionConflicts": [
"212 addresses appear on your newsletter suppression list. They are excluded here and should be excluded everywhere."
]
}Requirements
Integrations
No integrations required.
Before you start
- The segment and what you know about it
- Consent and suppression records
- Sequences already live
- Requires the Starter plan or higher
Use cases
Onboarding
A sequence built around what a new customer actually needs to do.
Re-engagement
Written for people who stopped opening — and checked before it runs.
Auditing what is live
Find the collisions and suppression gaps in sequences already running.
Pricing
Starter
Billed monthly through your PROMIVO subscription. Cancel at any time.
Runs consume your plan allowance for agent executions and tokens. See plan limits
FAQ
Is the consent check legal advice?
No. It reports what your own data records — a basis present, absent, or unclear — and flags segments to check. Marketing rules differ by country and it does not interpret them.
Will it send the emails?
No. It drafts and reports. Queuing a sequence puts your sending domain's reputation on the line, and that decision stays with a person.
Can it personalise deeply?
From what your data contains. It will not invent a company size, an industry or a use case to make a line land — a wrong personal detail is worse than none.
Reviews
Reviews open once customers have run this agent.
Need this agent to work differently?
Tell us what to change and our team will scope a customised version for your business.
Customize this agent