Email Sequence . Agency OS

What it does

Builds or rebuilds the always-on email flows that run behind your list: welcome, lead nurture, onboarding, re-engagement, cart abandonment and post-purchase. It designs the architecture first, the trigger that starts the flow, the spacing between sends and the exit conditions, then writes every email in full: subject line, preview text, body copy and a single CTA, each email doing exactly one job. Copy runs in your brand voice from your workspace files, objections from your audience get answered at the point in the sequence where trust has been earned, and every concrete claim is checked against your verified proof registry before it appears. The finished sequence is delivered as drafts plus a structured output that the critique, copywriting and conversion skills can read and improve.

Say this to start

This skill has no button. You start it by saying what you want. Any of these will do it:

> welcome sequence

> nurture flow

> drip campaign

> re-engagement emails

> onboarding email sequence

When to reach for it

leads opt in, get the lead magnet, and then hear nothing until you remember to email them
new customers or trial users churn before they ever reach their first win
a list has gone quiet and you want a structured way to wake it up without burning it
people start checkout and abandon it, and nothing follows them out the door
an existing flow underperforms and you want it rebuilt on sound structure instead of patched
a client engagement includes automated email flows as a deliverable

When NOT to use it

This skill builds evergreen flows only: sequences that run automatically, forever, for everyone who triggers them. It is not the tool for anything dated or one-off.

If you actually want	Use this instead
a one-off broadcast or newsletter issue written to the whole list on a given day	`newsletter-writer`
a dated launch calendar with an open, a close and launch-week sequencing	`legendary-copywriters` and its launch frameworks
a single high-stakes standalone email such as a cold pitch or a partnership ask	`legendary-copywriters`
the landing pages, forms and CRM automation the sequence plugs into	`funnel-architect`
chasing an unpaid invoice by email	`collections`

Before you start

What you need	Why
Brand DNA complete, or these supplied by hand: business name, what you sell and its price point, target audience, brand voice, email platform, core offer and your audience's key objections	every email is written in your voice toward your offer; the skill reads `voice-profile.md`, `positioning.md` and `audience.md` first and only asks for what the workspace cannot answer	Required
`verified-claims.md` populated with consented proof	any result, testimonial or named client that appears in an email must trace to the registry; anything that cannot gets flagged as a proof gap instead of published	Required
A clear trigger: what puts someone into this flow (form fill, purchase, signup, inactivity)	the first email references the exact action that started the sequence; a flow with a vague trigger reads generic from the opening line	Required
A decision on which workspace you are in, `_default` for your agency or `client--*` for a client deliverable	client-mode output carries the client's voice, offer and proof registry, never the agency's	Required
An ICA from the avatar skill	used for deeper personalization: which pains to open with and which objections to answer where; otherwise you are asked directly	Optional
Performance data from an existing sequence, if one is live	the optimization path depends on knowing where readers drop: opens, clicks or conversion	Optional

How it runs

Workspace and brand memoryDetects your workspace, loads voice-profile.md, positioning.md and audience.md under the freshness rules, and applies any learnings.md patterns tagged email-sequence that have recurred often enough to become rules.
AssessmentEstablishes the sequence type, the trigger event, what the reader already knows and believes at that moment, and the single conversion goal. Brand files answer most of this; you are only asked for the gaps.
Architecture before copySets the shape of the flow for the chosen type: how the emails ladder from value to ask, the spacing between sends, the exit conditions, and the subject line strategy, which is always clear over clever.
Write every emailEach email gets a hook, context, value and one CTA, in first person, in your voice, short sentences, signed with a name. One email, one job: nurture and sell never share a send.
Place the objectionsMaps your audience's objections to the position in the sequence where each is best answered, after value has been established and never in the opening email.
Claim gateChecks every concrete claim, result, testimonial and named client against verified-claims.md. Cleared claims appear exactly as cleared. Anything without a registry entry is flagged as a proof gap and genericized or omitted, even if you supplied the claim yourself.
Copy diagnosis gateScores each email on Attention, Resonance, Belief and Action in reading order, fixes the first failing axis before touching anything later, and holds the sequence until every axis clears the pass mark.
Delivery and loggingHands over the sequence overview, every email in full and the metrics plan, logs the deliverable to assets.md tagged email-sequence, then asks how the output landed and writes the answer to learnings.md so the next run starts sharper.

What you get

A sequence overview: name, trigger, goal, timing between sends and the exit conditions that take someone out of the flow
Every email written in full: subject line, preview text, body copy, send timing, the objection it handles and one CTA with button text and destination
A metrics plan naming the primary metric for this sequence and what to watch, with click rate treated as the reliable signal
A structured email-sequence-v1 output that downstream skills read: critique scores individual emails, legendary-copywriters rewrites or sharpens them, conversion-optimizer finds the funnel gaps around the sequence
An assets.md entry tagged email-sequence and a learnings log that calibrates future runs to your feedback

Honest limits

Read this before you rely on it

Claims in emails come only from your verified claims registry. An assertion without a registry entry, including one you state yourself in the brief, gets flagged as a proof gap and is not published until it is verified.
Nothing is sent by the skill. It produces drafts for your own email platform; connecting the flow, loading the drafts and pressing go are your job, on your account.
Evergreen only, by design. Dated launch mechanics, countdown timers and open-close windows are out of scope, and the skill will not fake urgency inside an always-on flow.
It writes the emails, not the plumbing. Forms, tags, automation triggers and the landing pages the CTAs point to sit with funnel-architect and your platform.
The copy gate sends weak drafts back before you ever see them, but plan for a revision pass on voice and emphasis rather than expecting first-draft-final.
Any benchmark it references is internal calibration, not a promise. Your list's own data is the only performance truth that matters.

Where people go wrong

The mistake	Do this instead
Stuffing company updates, blog links and a pitch into one email	One email, one job. If you need to nurture and sell, that is two emails on two days.
Piling CTAs into every send: book a call, download, watch, follow	One primary CTA, one button, one destination. A secondary action belongs in the P.S., not in a competing button.
Pitching in the first email of a welcome flow	Earn the right to sell. Value first, pitch after trust is built. The deeper the trust, the higher the conversion.
Subject lines that overpromise what the email delivers	Match subject intensity to content. A quick tip gets a quick tip subject line; save the bold claim for the email that delivers bold value.
A sequence that sounds like different people wrote different emails	Pick one voice and hold it. Read the whole flow back to back before it ships; if the sender changes personality mid-sequence, rewrite.
Ignoring the action that triggered the flow	Reference it in the first send: you grabbed this resource because of this problem. Context is what makes the rest of the sequence feel relevant.

Worth knowing

The sequences that convert do not read like campaigns. They read like a sharp friend who remembers exactly why you signed up and has something genuinely useful to say about it. Every rule in this skill, first person, one job per email, value before ask, exists to protect that feeling at automation scale.