mcpField
E
email-sequences

email-sequences

github.com/sanky369Marketing & growth3161 installsUpdated 16w ago
Repository

About email-sequences

A skill for planning and writing automated email sequences triggered by events such as signup, purchase, or launch. It serves marketers creating welcome, nurture, conversion, or launch campaigns, with a workflow that maps each email’s purpose and checks the sequence for consistency and truthful urgency.

Key features

  1. 01Selects sequence architecture by trigger
  2. 02Maps email timing and purpose
  3. 03Drafts subjects, previews, and email copy
  4. 04Checks calls to action and content quality
  5. 05Provides trigger, delay, and exit setup notes

Use cases

  1. 01Write a welcome series for new subscribers
  2. 02Nurture or convert prospective customers
  3. 03Plan emails around a product launch
  4. 04Set up automated emails triggered by purchases

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/sanky369/vibe-building-skills --skill email-sequences

After installing, type/email-sequencesin Claude Code to confirm it loaded.

Details

From the project README

Email Sequences

Write automated email sequences — fixed sets of emails sent on a schedule after a trigger (signup, purchase, launch date) — that build trust before they ask for anything. Prime directive: every email must be worth opening on its own, and every email must move the sequence's single job forward. A sequence with one job (deliver, nurture, convert, or launch) outperforms one trying to do everything; an email that only exists to "stay in touch" gets cut.

When to use / when not to

  • Use for any triggered/automated email series: welcome, nurture, conversion/sales, launch, or a custom variant of these.
  • If the user wants a recurring broadcast written fresh each time, that's skills/marketing/newsletter, not an automation.
  • If the user has no opt-in offer generating subscribers yet, run skills/marketing/lead-magnet first — a welcome sequence needs something to deliver.
  • If the user needs one high-stakes standalone sales page or email, skills/marketing/direct-response-copy goes deeper on a single asset.

Intake

Ask in one batch, only what's missing:

  1. Trigger and goal — what event starts the sequence, and what should the subscriber do/believe by the end?
  2. The offer (if any) — product, price point, main objections buyers raise.
  3. Audience temperature — cold strangers from an ad, warm lead-magnet subscribers, or existing customers? (This sets how early you can pitch.)
  4. Proof assets — testimonials, case studies, results, credentials available to use. Never invent these.
  5. Voice — sample emails or a brand-voice guide (use skills/marketing/brand-voice output if it exists).

Infer the sequence type from the trigger (Step 1 decision rule) rather than asking. If enough is known, state assumptions and draft — don't stall.

Workflow

1. Pick the sequence architecture

Decision rule — match on trigger:

TriggerSequenceEmails / spanJob
New subscriber (lead magnet, signup)Welcome5 over 7 daysDeliver promise, set expectations, earn trust, introduce offer softly
Subscriber finished welcome, no offer liveNurturerotating weekly patternStay valuable and top-of-mind between campaigns
Existing offer, warm listConversion7 over 14 daysPersuade a considered purchase
Something new going live on a dateLaunch7 over ~17 daysBuild anticipation, then open and close the cart

If the user's situation genuinely spans two (e.g., welcome that must sell a low-ticket offer), keep the primary architecture and borrow single emails from the other — never run two sequences at once to the same person.

2. Map the sequence

Lay out every email's day, job, and angle before writing a word. The canonical architectures (adjust days to the user's platform, keep the spacing logic — gaps grow as the sequence ages):

Welcome — Day 1: welcome + deliver the lead magnet + set expectations · Day 2: value + your story (why you do this) · Day 4: proof (case study or customer story) · Day 6: soft offer (introduce, no pressure) · Day 7: engagement ask (a genuine question that invites replies — replies also train inbox placement, a deliverability heuristic worth stating to the user).

Nurture — weekly, rotating pattern that repeats: story + lesson → framework or tool → case study/proof → soft offer → question/engagement. One value email between every ask, minimum.

Conversion — Day 1: name the problem + open a curiosity loop · Day 3: the mechanism (why common approaches fail, what's different about yours) · Day 5: the offer, plainly (what it is, who it's for, first CTA to buy) · Day 7: proof stack (testimonials, results) · Day 10: objection handling (take the top 2–3 objections from intake head-on) · Day 12: honest urgency (real deadline, real scarcity — see guardrail) · Day 14: final call (deadline restated, last CTA, no new arguments).

Launch — Day 1: announcement tease ("something's coming") · Day 3: the why + mechanism · Day 5: full reveal + benefits + early price/bonus · Day 7: early social proof · Day 10: objections · Day 14: urgency (closing soon / price rises) · Day 17: final call.

Propose the map as a table first if the situation is nonstandard; otherwise proceed straight to drafting.

3. Write every email

For each email in the map, in order:

  • Subject line: ≤50 characters, 2 candidates per email. Rotate styles across the sequence — curiosity ("the one thing X gets wrong about Y"), specificity ("how I [result] in [timeframe]"), benefit ("[benefit] without [drawback]"), question, urgency (only when urgency is real). No two consecutive emails with the same style.
  • Preview text: ≤90 characters extending, not repeating, the subject.
  • Hook (first 1–2 sentences): drop into a story mid-scene, a question, or a curiosity gap. Never open with "Hope you're doing well" or a recap of who you are.
  • Body: 100–200 words for value emails; sales emails (offer, objections, final call) may run to ~300 when the argument needs it. One idea per email. Short paragraphs. Written to one person.
  • CTA: exactly one per email, matched to the email's job — value emails end in a reply prompt or a single link; sales emails end in one buy/signup link (repeating the same link twice in a long email is fine; two different asks is not).
  • P.S.: optional; use for the soft-offer whisper in value emails or the deadline reminder in sales emails.

Thread the sequence: each email should reference or pay off something from the previous one (open loops close, stories continue) so the series reads as one narrative, not seven strangers.

4. Self-review and deliver

Run the quality bar per email and across the sequence. Deliver in the required output format with platform-agnostic setup notes (trigger, delays, exit conditions — e.g., "exit conversion sequence on purchase").

Required output format

# [Type] Sequence — [offer/audience]

## Sequence map
| # | Day | Job | Angle | CTA |
| 1 | 1 | ... | ... | ... |
(one row per email)

## Emails

### Email 1 — Day 1 — [job]
Subject A: "…" (n chars)
Subject B: "…" (n chars)
Preview: "…" (n chars)

[Full body, greeting through sign-off, CTA link marked as [LINK: destination]]

P.S. [if used]

---
(repeat for every email)

## Setup notes
- Trigger: [event] · Exit condition: [e.g., purchase removes from sequence]
- Delays: [as mapped] · Assumed platform capabilities: [merge name, delays, exit rules]

## Measure
- Watch: open rate by email (subject problem), click rate on offer emails (offer/copy problem), unsubscribe spikes (pacing problem), replies (health signal).
- Rule: rewrite any email whose opens fall far below its neighbors — it's the subject, not the list. (Heuristic.)

Quality bar

Per email — check, don't eyeball:

  • Subject ≤50 chars (count), makes a promise the body cashes, no clickbait mismatch
  • Preview ≤90 chars, not a subject repeat
  • Hook lands in the first 2 sentences; no "hope you're well" openers
  • One idea, one CTA; value emails 100–200 words, sales emails ≤~300
  • Merge fields used where natural ([FIRST_NAME]), never faked personalization ("I was just thinking about you")

Across the sequence:

  • Value-to-ask ratio: never two consecutive hard-sell emails except the final urgency → final call pair
  • Every objection from intake is answered somewhere before the final call
  • All proof (testimonials, numbers, results) supplied by the user or clearly marked [PLACEHOLDER — insert real testimonial]; never invented
  • Urgency is real (actual deadline/scarcity) or absent — fake countdowns are forbidden
  • Consecutive subjects use different styles; sequence reads as one continuing narrative
  • Send-day gaps follow the architecture (no two sales emails on back-to-back days)

Integration

  • skills/marketing/lead-magnet — supplies the opt-in promise the welcome sequence's Email 1 must deliver; pull the magnet's exact promise into intake.
  • skills/marketing/brand-voice — supplies the voice all emails match.
  • skills/marketing/direct-response-copy — deepens the offer/objection/final-call emails when the sale is high-ticket; hand it the offer and objection list.
  • skills/marketing/newsletter — subscribers exit the welcome/nurture sequence into the recurring newsletter; note the handoff in setup notes.
  • skills/marketing/content-atomizer — strong sequence emails can be atomized into social content, and vice versa.

Comments

Markdown supported