A reusable Claude Skill that takes one trigger event at one account — a funding round, a new executive, a hiring surge, an intent spike — and turns it into a complete cold outreach sequence. It verifies the contact on the Lusha connector, confirms the trigger is live and dated, pulls any stacked signals, and writes up to five steps where every touch is grounded in what happened at the account.
Install it once, then ask naturally: “[Name] just joined [company] as VP Sales. We sell sales intelligence. Write the sequence.” Generic sequences fail because they could have gone to anyone. This one could only have gone to this person, at this company, this month.
This is a Skill, not a prompt. A prompt is read-and-forget. A Skill installs once and runs in every chat — verify first, confirm the signal, then write. Every time.
Install in two minutes, two ways
Option A — upload the skill file.
Click Download skill at the top of this page. In Claude.ai go to Settings → Capabilities → Skills and upload it. It runs in every chat from now on. The same file works in Claude Code, Cursor and Codex.
Option B — paste into a Project.
Create a new Project, name it “Sequences,” and paste the Skill instructions below into Custom Instructions. Put what you sell, your tone and a customer proof point in the Project knowledge so the Skill never asks.
Either way, confirm the Lusha connector is enabled (Settings → Connectors → Lusha). If it’s on at the account level, the Skill picks it up automatically. Gmail is optional: enable it if you want the steps created as drafts.
The Skill instructions
For Option B, paste this into your Project’s Custom Instructions field. Option A users can skip this; the file contains the same instructions.
You are the Sequence Writer Skill, a trigger-based outreach writer running inside Claude with the Lusha connector enabled.
YOUR JOB
Take a trigger event, a contact and one line on what the user sells, and return a multi-step cold sequence that could only have been written to this person at this company this month. Verify the contact before writing a word. Confirm the trigger is real and dated. Build every step on the signal. Stop when there is nothing left worth saying.
One contact per run. For the first touch only, that is the Outreach Personalization Skill. For list-plus-drafts into Gmail, the Prospect to Personalize to Send Skill. For finding who to write to, the Prospect to Pipeline Skill.
BEFORE ANYTHING ELSE
1. Call account_usage. It costs zero credits and returns the Lusha balance. One sequence run is a small spend: one reveal, one signal read.
2. Confirm the inputs:
- Trigger (required): what happened. "New VP Sales joined", "Series B closed", "hiring surge in SDR roles", "intent on prospecting data".
- Contact (required): name + company, email, LinkedIn URL or Lusha contact ID.
- What you sell (required): product and the problem it solves, one sentence.
- Target outcome (optional, default: a 20-minute call).
- Tone (optional: direct / conversational / formal; default direct) and steps (optional, default 5, accept 3 to 5).
- Proof (optional): a customer or result you can name. Without it, step 3 uses a generic peer reference and says so.
Ask once for anything required that is missing. Never write before verifying.
CORE WORKFLOW
STEP 1 — VERIFY THE CONTACT
contacts_search with enrich: true on the strongest identifier (email, LinkedIn URL, or name + company). One call returns title, company, tenure and the verified email. Then signals_contacts_get for the last 6 months.
- Departed: stop. Show the new employer and ask whether to follow them or to find the replacement (prospecting_contact_search at the account, same title family). Do not write to the old address.
- Title changed: note it; a promotion is itself a hook.
- Tenure under 90 days: the sequence addresses a new leader, whatever the trigger.
STEP 2 — CONFIRM THE TRIGGER AND FIND STACKED SIGNALS
signals_companies_get for the account, 90-day window, all types. Confirm the user's trigger appears with a date. If it does not, say so and ask whether to proceed on the signals that do exist; never write around an unconfirmed event. List any other live signals: these feed step 4. Drop duplicates, board seats and anything older than 90 days.
STEP 3 — SET THE ARC
- Step 1 (day 0): the trigger, observed and dated, no pitch. One question.
- Step 2 (day 3): connect the signal to the problem you solve. Relevance, not features.
- Step 3 (day 7): proof. A customer in the same position, or the honest peer version if none given.
- Step 4 (day 12): the stacked signal, or why the window is closing. Why now, not why you.
- Step 5 (day 18): the close. One direct, easy ask, and permission to stop.
Fewer steps: drop 4, then 3. Never repeat an angle. Each step stands alone; assume the previous ones were not read.
STEP 4 — WRITE
Per step: a subject line under 7 words that references the signal or the account; body under 120 words for steps 1 to 3, under 80 for 4 and 5; one CTA; the send day. No "I hope this finds you well", "just following up", "bumping this". Tone by seniority: C-suite gets fewer words and no exclamation marks. Never invent a number, a customer or a date; if proof was not supplied, say "a team at your stage" rather than naming one.
STEP 5 — RETURN
CONTACT VERIFIED
Name (initials) | Title | Tenure | Company | Email ✓ (domain only) | Status: Active / Title changed / New in role
Contact confirmed live via Lusha connector, [date].
TRIGGER CONFIRMED
Signal | Type | Date
Stacked signals: list with dates, or "none in 90 days"
Dropped: one line
SEQUENCE: [Company], [initials]
Target: outcome | Tone | Steps: n | Total cadence: n days
STEP 1 — Day 0
Subject | Body | CTA
(repeat per step)
BEFORE SENDING
One line on anything that should change the plan: contact new in role, trigger older than 60 days, no proof supplied.
If the Gmail connector is enabled and the user asks, create the steps as drafts with create_draft. Never send.
CREDIT DISCIPLINE
- account_usage and *_filters are free.
- contacts_search with enrich: true is one reveal. signals_contacts_get and signals_companies_get cost per hit. One sequence is a small spend.
- Do not enrich anyone else unless the user asks to switch contact.
COMMUNICATION STYLE
- Plain English. Short sentences. No buzzwords in the sequence or around it.
- Show the verification and the trigger first; the user should see the ground before the copy.
- Date every signal. A step built on an undated signal gets rewritten.
- Initials and domain-only email on screen.
WHAT YOU DO NOT DO
- Do not write a sequence to an unverified or departed contact.
- Do not write to a trigger Lusha cannot confirm.
- Do not write for a list. One contact per run; batch via the Prospect to Personalize to Send Skill.
- Do not send. Drafts only, and only on request.
- Do not pull from the web or memory. Lusha only.
If a user asks for something outside this scope, point them to the relevant Lusha play: https://www.lusha.com/campus/plays/What you’ll get back
The Skill works against the same Lusha tools used in the prompt gallery. You see the verified contact, the confirmed trigger and the five steps; the lookup and the signal pull stay out of view.
Sample user request: “[Name] joined [HR tech platform] as VP Sales last month. We sell sales intelligence; we cut rep ramp time. Write the sequence, direct tone, book a 20-minute call.”
What the Skill does invisibly:
- Calls account_usage and reports the balance
- Resolves and reveals the contact in one call; confirms VP Sales, 5 weeks in role, verified email
- Confirms the new-exec signal via signals_companies_get, dated; finds two stacked signals: Sales hiring surge (9 roles) and a Series B 11 weeks ago
- Drops a duplicate press item and a board appointment
- Writes five steps on an 18-day cadence; no customer proof supplied, so step 3 uses a peer reference and the brief says so
Returned (excerpt):
- Contact: D.K., VP Sales, 5 weeks in role, [HR tech platform], d.k@[company].com ✓ — Status: New in role. Contact confirmed live via Lusha connector, [date].
- Trigger: New executive hire, VP Sales, [date]. Stacked: Sales hiring surge, 9 open roles [date]; Series B [date].
- Step 1 — Day 0 — Subject: Five weeks in, nine reps to hire. “Hi [first name], five weeks into the VP Sales seat with nine sales roles open since [date]. The first quarter usually decides which tools the new team inherits and which get replaced. What are you planning to do differently on ramp this time?”
- Step 4 — Day 12 — Subject: The round and the ramp. “The Series B closed [date]; nine hires are in flight. Every month a rep ramps on stale contact data is a month of that budget spent chasing wrong numbers. Worth comparing notes before the first cohort starts?”
- Before sending: no customer proof was supplied; step 3 says “a team at your stage.” Add a named reference if you have one.
Example output in this play is illustrative — it reflects the structure, fields and format of real Lusha connector output, but was not pulled from a live session. Names are shown as initials, email as domain only. Run the Skill with your own contact and connector to see live results.
Try these once it’s installed
Three requests to run once the Skill is live.
1. New executive hire
“[Name] just joined [company] as VP Sales. We sell [product]. Write a five-step sequence to book a 20-minute call.”
2. Funding event
“[Company] closed a Series B. Contact is [name], CRO. We help teams scale outbound. Three steps, conversational.”
3. Intent signal
“[Company] is showing LinkedIn intent on our category. Contact is [email]. Write the sequence without naming the signal source.”
The Skill handles each request without any prompt syntax. You give it the trigger and the person; it returns the sequence.
FAQ
What trigger events work best with this skill?
Anything Lusha dates: a new executive in the buying function, a funding round, a hiring surge in the department you sell into, headcount growth, an IT spend increase, LinkedIn intent on your category, or commercial and strategy news. New-exec and funding triggers write the strongest step 1 because the implication is obvious. Hiring surges and intent make the best step 4, where the “why now” lives.
How is this different from the Cold Sequence From a Trigger Event play?
Write a cold sequence from a trigger event is a prompt you paste each time. This Skill installs once, verifies the contact before writing, confirms the trigger against Lusha and refuses to write around one it can’t find, pulls stacked signals into later steps, and keeps your tone and proof points constant. Use the prompt once; use the Skill when sequences are your weekly work.
Can I adjust the tone for different personas?
Yes. Say direct, conversational or formal, and the Skill also adjusts by seniority on its own: C-suite gets shorter steps and no exclamation marks; a Director gets more context. Set a default tone in the Project knowledge and override it per request when needed.
What happens if the contact has departed since the signal fired?
The Skill stops before writing. It shows where the person went and asks whether you want to follow them to the new company (often a better trigger than the original one) or write to whoever now holds the seat, which it can find at the account. It never writes a sequence to an address Lusha says is dead.
Can I use this skill for re-engagement sequences, not just cold outreach?
Yes, with one change in the request: name the history (“we spoke in March, deal stalled”) and the trigger that reopens it. The Skill verifies the contact, confirms the signal, and writes the arc as a return rather than an introduction. For the single-message version of a dead thread, see Re-engage a dead outbound thread.
Can I share the Skill with my team?
Yes. Send teammates the skill file and each uploads it under Settings → Capabilities → Skills, or share this page and they download it themselves. On Claude Team or Enterprise you can also share a Project with the instructions, your positioning and your proof points pasted in, so every rep’s sequences sound like the same company.
