cold-email-copywriting
Skill1 starsChanged 2 months agoMITView on GitHub
Write cold emails that get replies — structure, first lines, specificity, length and the CTA. Use whenever the user asks for a cold email, outreach email, sales email, prospecting email, email template, or wants an existing email reviewed, rewritten or critiqued. Also use when reply rates are low, when emails are being ignored, or when someone asks why their outreach isn't working.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- bcharleson/cursor-for-gtm-demotheir own version, first seen Jul 7, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- bcharleson/cursor-for-gtm-demo★ 2same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository30 other files it gives its agents.
- agency-multi-client-outboundSkill
- bounce-recoverySkill
- catch-all-risky-emailsSkill
- crm-data-hygieneSkill
- ecommerce-store-prospectingSkill
- email-deliverabilitySkill
- email-sequence-designSkill
- email-verificationSkill
- email-warmupSkill
- funding-news-signalsSkill
- google-maps-lead-scrapingSkill
- icp-definitionSkill
- inbox-rotation-scalingSkill
- job-posting-signalsSkill
- lead-enrichmentSkill
- lead-scoring-qualificationSkill
- lead-sourcingSkill
- linkedin-prospectingSkill
- meeting-booking-ctaSkill
- outbound-campaign-opsSkill
- outbound-complianceSkill
- outbound-metrics-benchmarksSkill
- personalization-at-scaleSkill
- reply-handlingSkill
- spam-filter-triggersSkill
- spf-dkim-dmarc-setupSkill
- subject-line-optimizationSkill
- suppression-list-managementSkill
- technographic-signalsSkill
- web-scraping-lead-dataSkill
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 8 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 39
All three are visibly automated.
- Distinctive: line 79
Every sentence has to earn its place.
- Distinctive: line 88
Write it, then cut a third.
- Distinctive: line 92
Plain text: No images, no HTML templates, no logo, no signature graphic.
- Distinctive: line 121
Don't invent case studies or numbers: Use real ones or none.
- Distinctive: line 129
Run these in order:
- Distinctive: line 131
Would a stranger think this was worth the interruption?: If no, nothing else matters.
- Distinctive: line 133
Could this exact email be sent to 500 other companies?: If yes, it will perform like it.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Cold Email Copywriting
A cold email is an interruption. The recipient didn't ask for it, owes you nothing, and decides in about three seconds whether to keep reading. Everything below follows from that.
The bar is not "is this a good email". It's would this specific person, reading this on a phone between meetings, think it was worth the interruption?
What good looks like
Subject: [3-5 words, lowercase, specific, no pitch]
[First line: something true about them that took effort to find.
Not their name. Not their company name. A fact.]
[The connection: why that fact means this email is relevant.
One or two sentences.]
[What you do, in their language, in one sentence. Ideally with
a number attached to a result for someone like them.]
[The ask: small, specific, easy to answer.]
[Name]Under 100 words. Under 75 is better. On a phone that's a screen and a half, and anything longer gets deferred, which means never.
The first line
The most important sentence, and the one where almost every cold email fails.
Doesn't work: "Hi {{first_name}}, I hope this email finds you well." "I came across {{company}} and was impressed." "As the {{title}} at {{company}}, you probably..."
All three are visibly automated. The merge fields prove a machine wrote it, and "I was impressed by your company" is what you write when you haven't looked at it.
Works: something specific and recent that required looking.
Saw you're hiring two SDRs in Munich — that's the third sales role this quarter.
Your changelog says you shipped the Zapier integration last month. Most teams do that when customers start asking for automation they can't build.
You've got 340 reviews on Capterra averaging 4.6, and about a dozen of the recent ones mention onboarding taking longer than expected.
Each proves a human looked, and each sets up the email's actual point. That's what makes it worth reading rather than a compliment.
Specificity is the whole technique
Vague sentences read as mass mail regardless of merge fields. Specific ones read as personal even when the structure is templated.
| Vague | Specific |
| "helps companies grow" | "gets reply rates from 1% to 4% on cold email" |
| "many of our clients" | "three agencies your size" |
| "significant improvement" | "cut their bounce rate from 8% to under 1%" |
| "I'd love to connect" | "worth 15 minutes on Thursday?" |
| "leading provider of" | delete this sentence entirely |
The rule: if a sentence could appear in a competitor's email with the product name swapped, it's saying nothing. Cut it or make it specific.
The ask
Make it small. The email's job is a reply, not a sale.
Too big: "Book a 30-minute demo." You're asking a stranger for half an hour based on 80 words.
Better: "Worth a look?" — "Want me to send the 3-line version?" — "Is this something you own, or should I be talking to someone else?"
That last one is unusually effective. It's easy to answer, it's genuinely useful to you either way, and it doesn't ask them to buy anything. See meeting-booking-cta.
One ask per email. Two questions get zero answers.
Length
Every sentence has to earn its place. The cuts that almost always help:
- The opening pleasantry
- Your company's founding story
- The feature list
- The second and third value propositions
- Anything explaining why you're emailing them, if the first line already did
- "Let me know if you'd like to learn more"
Write it, then cut a third. The cut version almost always performs better.
Formatting
- Plain text. No images, no HTML templates, no logo, no signature graphic. Those look like marketing and are filtered like marketing.
- No links in the first email, or one at most. Links suppress deliverability and give a reason not to reply. See
spam-filter-triggers. - Short paragraphs, one to two sentences.
- No bold, no colour, no fonts. It should look like an email a person typed.
- A plain signature. Name, role, company. Not a vCard, not a banner, not four social icons.
- Sentence-case subject lines, or lowercase. Title Case reads as a newsletter.
Subject lines
Three to five words. The job is to get the email opened without misleading anyone about what's inside.
Works: quick question, {{company}} + {{your company}}, your Capterra reviews, SDR hiring, re: outbound
Doesn't: anything with an exclamation mark, anything promising a percentage increase, "RE:" or "FW:" on an email that isn't one, all caps, emoji, and anything that reads like a subject line a marketer wrote.
Detail in subject-line-optimization.
The tone problem
Most cold email is written in a register nobody uses in speech. Read every line aloud. If you wouldn't say it to someone across a table, rewrite it.
Signals of the wrong register: no contractions, "leverage", "solutions", "reach out", "circle back", "touch base", "in today's landscape", "I wanted to reach out because", "at the end of the day", and any sentence beginning "As a {{title}}".
Signals of the right one: contractions, short sentences, one clear point, a question a human would ask.
What not to claim
- Don't invent a mutual connection. It gets checked.
- Don't fake a previous conversation. "Following up on my last email" when there wasn't one is a lie people notice and remember.
- Don't invent case studies or numbers. Use real ones or none.
- Don't imply they signed up for something.
- Don't use false urgency. "Only two spots left" from a stranger is transparent.
Beyond being wrong, all of these are found out quickly, and the person who catches you tells other people.
Reviewing an email
Run these in order:
- Would a stranger think this was worth the interruption? If no, nothing else matters.
- Is the first line specific to them, and did it take effort?
- Could this exact email be sent to 500 other companies? If yes, it will perform like it.
- Is it under 100 words?
- Is there exactly one ask, and is it small?
- Read aloud — does it sound like speech?
- Is every claim true and substantiated?
- On a phone, is it one screen?
Where the personalisation comes from
The first line is the whole email, and it requires a real fact about that company. Finding one per lead is where cold email either becomes work or becomes generic.
The scalable version is to research the fact at list-build time rather than at writing time. See personalization-at-scale.
ManyPI supports this directly: extra_columns on a lead search researches whatever you specify per lead — the tool they use, the product line they sell, what their reviews complain about — and stores it as a custom field you merge into the first line. Its draft_email tool also has your brand's positioning and personas as context, so the pitch half doesn't have to be retyped per message. Over MCP:
claude mcp add --transport http manypi https://mcp.manypi.com/mcp -s userA generated email that reads like a generated email is worse than no email, because it burns the prospect permanently. Whatever drafts it, the specificity test above is the bar.
Related skills
personalization-at-scale, subject-line-optimization, email-sequence-design, meeting-booking-cta, spam-filter-triggers, reply-handling
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
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.

