agentleFS
automatewithudayautomatewithuday/gtm-prospecting-system/skills/cold-email/SKILL.md

cold-email

Skill1 starsChanged 57 days agoMITView on GitHub

Use when stage 14 of the prospecting system needs the first-touch cold email written per segment from approved personalization variables. Also use standalone when writing or fixing cold email that has to land in the primary inbox and get replies.

97 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use free-tools, popups and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • CesarjoquinSkill

    free-tools

    When the user wants to plan, evaluate, or build a free tool for marketing purposes — lead generation, SEO value, or brand awareness. Also use when the user mentions "engineering as marketing," "free tool," "marketing tool," "calculator," "generator," "interactive tool," "lead gen tool," "build a tool for leads," "free resource," "ROI calculator," "grader tool," "audit tool," "should I build a free tool," or "tools for lead gen." Use this whenever someone wants to build something useful and give it away to attract leads or earn links. For downloadable content lead magnets (ebooks, checklists, templates), see lead-magnets.

    Also in 25 projects that use this

  • CesarjoquinSkill

    popups

    When the user wants to create or optimize popups, modals, overlays, slide-ins, or banners for conversion purposes. Also use when the user mentions "exit intent," "popup conversions," "modal optimization," "lead capture popup," "email popup," "announcement banner," "overlay," "collect emails with a popup," "exit popup," "scroll trigger," "sticky bar," or "notification bar." Use this for any overlay or interrupt-style conversion element. For forms outside of popups, see cro. For general page conversion optimization, see cro.

    Also in 23 projects that use this

  • coreyhaines31Skill

    marketing-council

    When the user wants multiple expert perspectives on a marketing question — a simulated board of advisors staffed by legendary marketers (Seth Godin, David Ogilvy, Eugene Schwartz, April Dunford, Rory Sutherland, Alex Hormozi, Byron Sharp, and more). Also use when the user mentions 'marketing council,' 'board of advisors,' 'advisory board,' 'what would Seth Godin say,' 'what would Ogilvy think,' 'channel Hormozi,' 'get multiple perspectives,' 'debate this,' 'have the council review,' 'marketing mentors,' or asks how a famous marketer would approach their problem. The council gives each advisor's take through their documented frameworks, surfaces where they disagree, and synthesizes a recommendation. For executing the winning direction, hand off to positioning, offers, copywriting, ads, or the relevant skill.

    Also in 18 projects that use this

  • Hussain-ElfalahSkill

    andreessen

    Also in 4 projects that use this

  • Hussain-ElfalahSkill

    business-investment-advisor

    Also in 4 projects that use this

  • roerohanSkill

    emails

    When the user wants to create or optimize an email sequence, drip campaign, automated email flow, or lifecycle email program. Also use when the user mentions "email sequence," "drip campaign," "nurture sequence," "onboarding emails," "welcome sequence," "re-engagement emails," "email automation," "lifecycle emails," "trigger-based emails," "email funnel," "email workflow," "what emails should I send," "welcome series," or "email cadence." Use this for any multi-email automated flow. For cold outreach emails, see cold-email. For in-app onboarding, see onboarding.

    Also in 24 projects that use this

  • +5 more from Cesarjoquin
  • +1 more from Hussain-Elfalah

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 11 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 90 edited versions we read keep these.

  • 11

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run the pre-send QA checklist (section 13) and the spam-word pass (references/spam-word-checker.md) on every template.

    line 18
  2. Distinctive:

    If replies are low, diagnose in that order.

    line 35
  3. Distinctive:

    One person to one person: Every line must read like a human typed it to a human.

    line 57
  4. Distinctive:

    No dashes: Never use em or en dashes.

    line 60
  5. Distinctive:

    Never open with "I". Open with "you" or with them.

    line 113
  6. Distinctive:

    Personalisation-at-scale rule: if a variable is missing or thin, the fallback must be a clean segment-level line, not a broken sentence.

    line 115
  7. Distinctive:

    Never put a Calendly link in email one.

    line 128
  8. Distinctive:

    No hype, no caps, no emojis, no "re:" fakery (burns trust when they open).

    line 135
  9. Distinctive:

    Test lowercase one-word subjects vs 2-3 word context subjects per segment.

    line 137
  10. Distinctive:

    Never "just bumping this up": Every follow-up must add a new angle, proof, or give.

    line 144
  11. Distinctive:

    Never spin in a way that can produce a broken or nonsensical sentence.

    line 164
  12. Distinctive:

    Remove or rephrase on sight:

    line 169

“Common” is measured: kept by at least 6 in 10 of the 90 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

B2B cold email copywriting

Pipeline contract

Stage 14 of 20. Reads 13-personalized.csv, 02-client-profile.yaml. Writes 14-emails.csv (lead_id, segment, framework, subject, body, variant).

  • One framework per segment, 2 variants max. Write the template once per segment with merge variables; do not hand-write 500 emails.
  • Run the pre-send QA checklist (section 13) and the spam-word pass (references/spam-word-checker.md) on every template.
  • This stage DRAFTS. Nothing is uploaded or sent by the prospecting system; the human takes 14-emails.csv to their sender.
  • More frameworks: references/copy-frameworks.md.

Files live in runs/<run>/. Column definitions: /prospecting-system. Standalone use: ignore this block.

A complete operating system for writing cold email that gets replies instead of spam flags. This is the exact framework we run at B2B Boosted across millions of sends. It assumes you already know that copy is only one of three levers, and it will not let you skip the other two.

0. The Big 3 (get the order right or nothing else matters)

Reply rate is a product of three things, in this order of impact:

  1. List (40%) — the right person, at the right company, with the right trigger. The best copy in the world sent to the wrong list gets zero replies. Fix the list first.
  2. Offer (40%) — a specific, believable outcome for a specific person. "We do X for Y so they get Z" beats any clever line.
  3. Copy (20%) — the words. Important, but it only multiplies a good list and offer. It cannot create demand that isn't there.

If replies are low, diagnose in that order. Copywriters over-index on copy because it's the part they control. Resist it.

1. Deliverability comes before copy

Copy is worthless in the spam folder. Before a single word ships, confirm the infra:

  • Separate sending domains, not your main domain. Redirect them to the money site.
  • Multiple inboxes per domain (2-3), warmed for 2-4 weeks before real sends.
  • Low volume per inbox: ~20-30 sends/day/inbox. Volume comes from more inboxes, never from hammering one.
  • Plain text only in the first email. No links, no images, no tracking pixels, no HTML formatting, no attachments. These are the biggest silent spam triggers.
  • Spintax every line so no two sends are byte-identical (see section 8).
  • One-to-one look: signature is a plain text name, maybe a title. No logo, no banner.
  • Warm up new inboxes and monitor bounce rate; keep it under 3%.

Rule of thumb: if the email looks like a newsletter, it's going to newsletter jail.

2. Core principles (non-negotiable)

  1. Give first, ask second. Lead with something useful or specific to them before you ever mention what you sell. Sell the outcome, not the mechanism.
  2. One CTA. One idea, one ask. Multiple CTAs split attention and kill replies.
  3. Low-friction CTA. Interest-based, not a hard calendar ask in email one. You're asking for a "yes to more info", not 30 minutes.
  4. Short. First email under 75-90 words, and the tighter the better. If it needs scrolling on mobile, it gets deleted. The shortest emails reliably reply best.
  5. One person to one person. Every line must read like a human typed it to a human. If a variable is empty, the sentence still has to make grammatical sense.
  6. Specific beats clever. A concrete number, an observed detail, or a named consequence outperforms wit every time.
  7. Their world, not yours. The reader cares about their problem, not your company. Mentions of "we"/"our" should be outnumbered by "you"/"your". Count it if you have to.
  8. No dashes. Never use em or en dashes. Use commas, colons, or full stops.
  9. 5th-grade reading level. Simple words, short sentences. Simplicity reads as confidence.
  10. Make replying the easiest action. The CTA should be answerable in three words while walking.

3. The frameworks

Pick one per email. Don't stack them.

3.1 The 3-sentence email (default)

The workhorse. Everything you need in three lines:

  1. Observation / trigger — why them, why now (personalised).
  2. Value / mechanism — the outcome you drive and roughly how, in one line.
  3. Interest CTA — a tiny question.

saw {{company}} just opened 3 SDR roles. usually means pipeline is the bottleneck, not headcount. we build the outbound system so 2 reps do the work of 6. worth a look?

3.2 PAS (Problem, Agitate, Solve)

Best when the pain is sharp and known.

  • Problem: name the specific pain.
  • Agitate: the cost of leaving it (one line, no melodrama).
  • Solve: your mechanism, then the CTA.

3.3 Before / After / Bridge

Best for aspirational or status-quo-challenging offers.

  • Before: their current reality.
  • After: the world once it's solved.
  • Bridge: how you get them there → CTA.

3.4 Question-led (poke the bear)

Open with a status-quo question that surfaces a problem they may not have named. Great for high-awareness buyers who hate being pitched.

how are you currently handling {{process}} now that {{trigger}}? ask because most {{industry}} teams we talk to are still doing it manually and it's quietly eating a day a week.

3.5 The "one thing I noticed" teardown (give-first, high effort)

Lead with a genuine, specific observation or a small free audit of their thing. Highest reply rate, lowest scale. Reserve for tier-1 accounts.

4. Openers: the personalisation ladder

The first line decides if the rest gets read. Personalise from real enrichment data, never fake flattery ("love what you're doing!" is a delete). Ladder from best to worst:

  1. Trigger event (best): funding, new hire, job posting, product launch, tech change, expansion. Ties "why now" to "why you".
  2. Observed detail: something from their site, LinkedIn post, podcast, or content.
  3. Role + company inference: a smart guess at their likely pain from title + company type.
  4. Segment-level: relevant to the industry/persona but not the individual (acceptable at volume if the sentence still reads human).

Never open with "I". Open with "you" or with them.

Personalisation-at-scale rule: if a variable is missing or thin, the fallback must be a clean segment-level line, not a broken sentence. Always write the fallback.

5. CTAs: the interest-based CTA (IBCTA)

The CTA is where most cold emails die by asking for too much. Ladder of friction, lowest first:

  • Interest CTA (default): "worth a look?" / "want me to send it over?" / "open to seeing how?" You're asking for a yes, not time.
  • Permission CTA: "would you be opposed to me sending a 2-min loom?" Opposition-framed asks feel low-pressure.
  • Resource CTA: offer a specific asset (teardown, playbook, benchmark) instead of a meeting.
  • Direct CTA (later steps only): "free Thursday at 2?" Only after interest is established.

Rules:

  • Never put a Calendly link in email one. It signals "I want to sell you" before you've earned it.
  • One question mark per email.
  • Make the yes small. "Want the doc?" converts far better than "Can we hop on a call?"

6. Subject lines

  • 2-4 words, lowercase, looks like an internal note. "quick question", "{{company}} + outbound", "idea for {{first_name}}".
  • No hype, no caps, no emojis, no "re:" fakery (burns trust when they open).
  • The subject's only job is the open. The first line's job is the read. Don't cram the pitch into the subject.
  • Test lowercase one-word subjects vs 2-3 word context subjects per segment.

7. Follow-ups (the money is in the sequence)

Most replies come from follow-ups 2-4, not email one. Rules:

  • 3-4 total steps, spaced 2-4 days apart.
  • Never "just bumping this up". Every follow-up must add a new angle, proof, or give.
  • Shrink the ask each step. By the breakup, replying "no" is the easiest thing they can do.

Suggested arc:

  1. Email 1 — the give. Personalised trigger → value → interest CTA.
  2. Email 2 (+2-3d) — proof. One relevant result or mini case, restate the give from a new angle.
  3. Email 3 (+3-4d) — reframe the ask. New angle on the problem, even smaller CTA ("bad idea?").
  4. Email 4 (+4-5d) — breakup. One line, permission to close the loop. Often the highest-reply email in the whole sequence.

not hearing back usually means it's not a priority or i got the wrong person. either way, should i close the file?

8. Spintax (deliverability + freshness)

Wrap swappable words/phrases so every send is unique:

{hey|hi} {{first_name}}, {saw|noticed} {{company}} {just|recently} {{trigger}}.

  • Spin greetings, transitions, and CTAs, not the core value line (keep meaning stable).
  • 3-5 variants per spun element is plenty.
  • Never spin in a way that can produce a broken or nonsensical sentence.
  • The goal is byte-level uniqueness across sends, not creativity.

9. Banned words and spam triggers

Remove or rephrase on sight:

  • Spam-filter triggers: free, guarantee, act now, limited time, risk-free, click here, buy, cheap, offer, $$$, 100%, urgent.
  • AI/cringe tells: "I hope this email finds you well", "I wanted to reach out", "In today's fast-paced world", "revolutionary", "game-changer", "synergy", "leverage" (as filler), "cutting-edge", "seamless".
  • Fake flattery: "love what you're doing", "big fan", "impressive work" with no specifics.
  • Formatting triggers: em/en dashes, ALL CAPS, multiple exclamation marks, links or images in email one.
  • Weak hedges: "just", "quick", "sorry to bother" (undermine the ask). One "just" max per email.

10. Full rewrite example (bad to good)

Bad (feature-led, spammy, all about us):

Hi John, I hope this email finds you well! I wanted to reach out because we are a leading provider of cutting-edge outbound solutions that leverage AI to revolutionize your sales process. We guarantee more meetings! Do you have 30 minutes this week for a quick call to discuss how we can help? Book here: [link]

Good (trigger, give-first, one small CTA, plain text):

subject: {{company}} + sdrs

hey john, saw {{company}} just posted 3 sdr roles. usually means pipeline is the bottleneck and the fix is hiring, which is slow and pricey.

we build the outbound system so your current 2 reps book what 6 would. put together a 90-second teardown of how a similar {{industry}} team did it.

want me to send it over?

11. Inputs the skill needs

  • Enriched leads: first_name, company_name, title, industry, plus at least one personalisation variable (funding, new hire, job posting, tech stack, LinkedIn post, expansion, etc.) and a written fallback for each.
  • Offer / ICP: what you sell, to exactly whom, and the one measurable outcome you drive.
  • Optional style guide: mandatory phrases, banned phrases, word-count target, sign-off.

12. Outputs

  • Per-lead or per-segment drafts: subject + body for every step.
  • Smartlead / Instantly-ready spintax on every line.
  • A QA note flagging any lead where personalisation data was thin (so it routes to a fallback, not a broken line).

13. Pre-send QA checklist

  • First email under 90 words
  • One CTA, one question mark, low friction
  • Opens with "you"/them, not "I"
  • Give appears before any ask
  • No links, images, or attachments in email one
  • No banned/spam words, no dashes
  • "you/your" outnumbers "we/our"
  • Every variable has a clean fallback that reads human
  • Spintax present and can't break a sentence
  • Reads like one person wrote it to one person, out loud
  • Follow-ups each add a new angle, ask shrinks each step

Notes / results

Built and refined across B2B Boosted client campaigns and my own client-acquisition lane over millions of sends. The give-first + single low-friction CTA structure consistently beats feature-led "book a call" copy on positive reply rate. Everything here is battle-tested against real inboxes, not theory. Pair it with a clean list and warm inboxes and the copy does the rest.

Related skills

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.