cold-email-writer
Skill0 starsChanged 40 days agoMITView on GitHub
Write, rewrite, and critique cold outreach emails that get replies. Covers investor/fundraising outreach, customer and sales prospecting, partnership asks, recruiting, press and advisor intros, and follow-ups. Use this skill whenever the user asks for a cold email, cold DM, outreach message, investor email, intro request, "email to a stranger", follow-up to an ignored email, or wants a draft reviewed or tightened, even if they never say "cold email". Also use it when they paste a draft and ask why it isn't getting replies. Drafts are short, proof-first, and free of fake personalization.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- prosyo-ai/gtm-skilltheir own version, first seen Oct 25, 2026
- koinod/skillstheir own version, first seen May 22, 2026
- koinod/koino-skillstheir own version, first seen May 22, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- prosyo-ai/gtm-skill★ 13same name
- koinod/koino-skills★ 7same name
- koinod/skills★ 0same 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 repository5 other files it gives its agents.
- AGENTS.mdAGENTS.md
- fewwords-summarizerSkill
- human-pencil-detSkill
- human-pencilSkill
- musk-email-writerSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 4 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
4
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Write cold emails a busy stranger will actually answer.
- Very specific: line 16
Decide on personalization: Read references/personalization.md.
- Distinctive: line 21
If the user pastes a draft, keep their facts, diagnose in 3-5 bullets (what reads as template, what's buried, what's unverifiable), then …
- Distinctive: line 36
Every claim checkable? No "proven 10x ROI" or "$18B market" without a source or definition.
- Very specific: line 47
Add the follow-up (see references/followups-and-delivery.md) only if asked or if the user is sending a sequence.
- Distinctive: line 49
Keep commentary tight. Don't lecture about the rules unless asked.
- Very specific: line 57
Default: plain, direct, founder-voice.
“Common” is measured: kept by at least 6 in 10 of the 3 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.
Files
Cold Email Writer
Write cold emails a busy stranger will actually answer. Distilled from Y Combinator's cold-email advice, OpenVC's critique of the "custom hook" template, a 5-rule founder slide, Alex Hormozi's offer and sales advice, and the terse style of ~200 documented Musk emails.
Core idea: a cold email earns a reply with proof and clarity, not flattery or fake personalization. AI and mail-merge made the "Heard you on the podcast, congrats on X, since you're big on Y..." hook free, so recipients now read it as a template.
Workflow
- Gather inputs. Need: sender (founder or SDR/agent?), recipient, what you do, best proof, the ask. If something is missing, draft anyway with
[placeholders]and list what to fill in. Ask at most one question, only if the draft can't be written sensibly without it. 1b. Check the offer (sales, services, agency outreach). If there is no concrete offer (outcome, mechanism, timeline), fix that first. Readreferences/offer-and-sales.md. - Decide on personalization. Read
references/personalization.md. Default is no hook. Include one only if it passes the "could this be pasted onto another recipient?" test. - Draft using the structure below. Read
references/principles.mdfor the full rule set andreferences/examples.mdto calibrate length and tone. - Run the self-check below and fix failures silently.
- Deliver in the output format.
If the user pastes a draft, keep their facts, diagnose in 3-5 bullets (what reads as template, what's buried, what's unverifiable), then give the rewrite.
Structure (3-5 sentences, under 60 seconds to read)
- What you do, plain words, zero jargon. Say what the product does, not the category.
- Best proof, one or two checkable facts: paid pilots, revenue, growth, launch, named customers, a technical team with relevant credentials. Put the strongest first.
- Why it matters to them (only if real and specific). Their pain in their words, or one genuine reason it's them. For sales, lead with the outcome and name the pain accurately (see
references/offer-and-sales.md). - The ask, explicit and on its own line right before the sign-off. Ask for something small: a reply, advice, an intro, permission to send a short demo or deck. Not a 30-minute meeting from a stranger.
- Sign off with name and role. Short, casual subject line (what + proof, e.g. "Automated bank QA, 2 paid pilots").
Self-check before delivering
- Under ~100 words? Readable aloud in under a minute?
- First sentence says what the company does, with no warm-up?
- Zero flattery, zero backstory, zero origin story?
- Every claim checkable? No "proven 10x ROI" or "$18B market" without a source or definition.
- Any hook passes the swap test, or is removed?
- One goal, one ask, on its own line, smaller than a meeting?
- Sounds like a person talking, not an AE or a pitch deck? ("Cheers", "I'd love to pick your brain", "game-changing" = cut.)
- Subject line short and casual?
- (Sales) Is there a real offer: outcome, mechanism, timeline? Scarcity or guarantee only if true?
Output format
- The email (subject + body) in a block.
- Under it, 2-4 lines: what's placeholder, what claim needs a source, and one optional variant (e.g. shorter, or with a real hook) only if useful.
- Add the follow-up (see
references/followups-and-delivery.md) only if asked or if the user is sending a sequence.
Keep commentary tight. Don't lecture about the rules unless asked.
Sender matters
Founder-sent emails get away with more brevity because the sender is the proof. An SDR, agent, or AI sender has to earn the reply from the email alone: stronger proof, clearer fit, borrowed credibility (customer names, YC, past employers). Ask or infer who is sending and adjust.
Tone
Default: plain, direct, founder-voice. If the user wants it blunter or more "Musk-style", read references/musk-tone.md. This is style only. Never present output as a real Musk email, and note that his emails are mostly warm or internal, so brevity there leaned on existing credibility.
Reference files
references/principles.md: the consolidated rule set and where sources agree or conflictreferences/personalization.md: when and how to personalize, the swap testreferences/examples.md: before/after rewrites (investor, sales) and a bad-vs-good checklistreferences/offer-and-sales.md: Hormozi offer rules mapped onto email copy, value equation, sales-process rules, how they reconcile with YC/OpenVCreferences/musk-tone.md: terse-style moves and their limits for cold usereferences/followups-and-delivery.md: follow-up cadence, ignored emails, SPF/DKIM/DMARC check
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.

