Are these really free, and what is the catch?

Free, and every word is on this site to read before you take it. The catch, if it is one, is that they encode our opinions about how outbound should be run. We think that is the reason to use them rather than a reason not to, and where an opinion is doing the work the file says so.

Did you write all of these?

No, and each page says which. 2 are ours from the first line: the capacity planner and the PECR checker. The other 9 are adaptations of MIT-licensed skills published by other people: eight from open-source repositories and one from Corey Haines. We gathered them, took the vendor references out, rewrote them in British English, and put our own rules in where the original was silent or where we disagree. The licence and a list of what changed are in the frontmatter of every adapted file, in a Credits section at the bottom of it, and in a provenance table on its page. The zip carries the licence terms in its LICENSE.

Why these ones and not the whole of the source library?

Because we only publish what we would run ourselves. The upstream set has thirty-nine files. The ones we left out are walkthroughs of one vendor's interface, skills that need a call-recording tool, or things that are not outbound at all. A library padded to forty would rank about the same and would waste the time of anyone who trusted it.

Do I have to give you an email address?

No. There is no form on this page. Each file is displayed in full with a copy button and a download button, and the zip is a plain link, so an address would buy you nothing you cannot already have.

Can I change them, or use them with clients?

Yes to both. Copy them, edit them, put them in front of your own clients. That is what the MIT licence on them is for, and it is the same licence that let us adapt the ones we adapted. If you improve one we would like to hear about it, but nothing is conditional on that.

Are you affiliated with Anthropic?

No. Claude is Anthropic's product. These are files written by an outbound agency in Middlesbrough that happen to be in the format Claude reads, and nothing here is endorsed by or connected to Anthropic.

↗Free · nothing gated

11 skills
for the jobs outbound is made of.

Not a landing page in front of a PDF. Each file is on this site in full, with a copy button and a download button, because a skill you cannot read before installing is one you should not install.

Download all 11 (.zip)Browse by job

No email address, no account, nothing to sign

These are working files. Each one takes a job we do by hand, writes down how it is actually decided, and hands the whole thing over. 2 of them we wrote from scratch. The rest we gathered from open-source skill libraries, cut the vendor references out of, rewrote in British English, and folded our own rules into: approved lists, the UK legal-form split, link-free first emails, sequence-length arithmetic. The original repository and licence are named on every page, because the method is not the moat and neither is pretending we invented all of it.

Everything in one zip

All 11 folders, ready to drop into ~/.claude/skills/, plus a README with the credits table and a LICENSE carrying every upstream author's notice. Rebuilt from the committed files on every deploy, so the zip and the pages cannot drift apart.

Download all 11 skills

.zip · 50 KB · no email address

The library

11 skills, grouped by the job they do

The line under each card is the description your assistant matches against, taken from the file rather than written for the page. Where a skill began as somebody else's work the card says so, and its page says exactly what changed.

Targeting

Who to contact and why now, decided before a single row is loaded.

trigger-finderAdapted from open source, MIT

A Claude skill that turns buying signals into timed outreach

Which signals justify contact now, which need a human first, and what to say either way.

Used when: Finds and interprets buying signals so outreach lands at the moment something changed. Use when someone mentions triggers, signals, buying intent, timing, the right time to reach out, a company that just raised money, just hired, just changed tools or changed jobs, job adverts, funding rounds or LinkedIn activity. Use even when the question is only when should I contact this company. Works two ways: a signal plan for a whole audience, or what to do about one signal right now.

Read the whole file Download .md

Offer and angle

What you are actually proposing, and the one line that has to carry it.

cta-designerAdapted from open source, MIT

A Claude skill that writes the ask at the end of a cold email

Five asks that are not a calendar link, and the rule that keeps them out of a first email.

Used when: Designs the ask at the end of a cold email or LinkedIn message so it gives something away instead of demanding a meeting. Use when asked what call to action to use, how to end an outreach message, what to offer a prospect, how to get a reply, or what to ask for at each step of a sequence. Use before finalising any outreach copy, because the ask decides whether the rest of the message gets a reply.

Read the whole file Download .md

offer-definerAdapted from open source, MIT

A Claude skill that turns features into an offer

One line, one paragraph, one page, and the rule that keeps invented numbers out of all three.

Used when: Turns a product description or a list of features into an offer somebody would reply to, at three lengths. Use when asked how to frame an offer, what to say in a cold email, how to pitch a product, how to explain what we do, how to turn features into outcomes, or for a one-liner or value proposition. Use after the audience is decided and before any outreach copy is written, because copy cannot rescue an offer nobody wants.

Read the whole file Download .md

Writing and replies

The emails themselves, and what to do when somebody answers.

copywriting-refinerAdapted from open source, MIT

A Claude skill that audits cold email copy and rewrites what fails

Eleven checks on a cold email, every failure quoted, and the rewrite in the same reply.

Used when: Audits a cold email, a LinkedIn message or a whole sequence against a fixed checklist, then rewrites every element that fails. Use when someone asks to refine, review, tighten or fix outreach copy, pastes an email and asks whether it is any good, too long, too formal or too salesy, or shares a sequence before it goes out. Always returns a pass or fail audit and a corrected version, never vague feedback.

Read the whole file Download .md

cold-email-writerAdapted from Corey Haines, MIT

A Claude skill that writes the cold email itself

The whole email, written to the limits, with the frameworks stripped of invented results.

Used when: Writes B2B cold emails and short sequences that read as though a person wrote them. Use when someone wants a cold email, a prospecting or outbound email, an SDR email, a campaign, or says nobody is replying to their emails. Covers subject lines, opening lines, the body, the ask, personalisation and follow ups. Use before sending anything cold, and before rewriting a campaign that is getting no answers.

Read the whole file Download .md

copywriting-first-touchAdapted from open source, MIT

A Claude skill that writes the first email in a sequence

The opener, calibrated to seniority, with the three seconds it actually gets.

Used when: Writes the first email to a cold prospect, either on its own or as the opener of a sequence. Use when someone asks for a first email, a cold email draft, a first touch, how to open with a particular prospect, or an outreach email with no prior context. Produces one email with a subject line, a body inside the word limit, the reasoning behind each choice, and the setup for the next two emails when it opens a sequence.

Read the whole file Download .md

copywriting-follow-upAdapted from open source, MIT

A Claude skill that writes follow-ups worth sending

What each follow-up is for, and when to stop rather than send another.

Used when: Writes the follow-up emails after no reply, each one carrying a new angle rather than a reminder. Use when someone asks for a follow-up, a second or third email, a bump, a breakup email, or what to send when a prospect has gone quiet. Use before sending anything that begins with just checking in, because the position in the sequence decides the angle and this file sets both.

Read the whole file Download .md

reply-handlerAdapted from open source, MIT

A Claude skill for the replies that are not answers

Seven categories, one decision each, so the inbox is not re-decided at nine at night.

Used when: Classifies a reply to cold outreach and writes the response, including the replies that are not answers at all. Use when someone shares a prospect's reply and asks how to respond, what to say to an objection, how to handle an out of office, a wrong-person redirect, a request to be removed, a pricing question, or a maybe later. Use before writing any reply to a prospect, so the category decides the response rather than the mood of whoever is in the inbox.

Read the whole file Download .md

Measuring, capacity and compliance

Whether the numbers are any good, how many leads a day you can sustain, and who you may lawfully email.

inbox-capacity-plannerWritten by TheCold.Ai

A Claude skill that computes your daily new-lead cap

How many new leads a day your mailboxes can sustain, and what breaks when you exceed it.

Used when: Works out how many new leads a day a set of sending mailboxes can actually sustain, given how long the sequence is, and flags when a planned import will starve its own follow-ups. Use when sizing a mailbox fleet, planning cold email volume, or diagnosing a campaign that looks healthy on the dashboard but has stopped reaching anyone new.

Read the whole file Download .md

uk-pecr-cold-email-checkerWritten by TheCold.Ai

A Claude skill that checks a UK cold email list against PECR

Which rows on your UK list you are actually allowed to email, with the regulation for each answer.

Used when: Screens a UK B2B prospect list and a draft cold email against PECR and the UK GDPR before anything is sent. Classifies each recipient by legal form, because that is what decides whether you may email them at all, and checks the message itself against the sender-identity and opt-out rules. Use before loading a list, or when auditing a campaign that is already running.

Read the whole file Download .md

outbound-analystAdapted from open source, MIT

A Claude skill that tells you whether your outbound numbers are good

A straight verdict on a reply rate, with the reference range and who measured it.

Used when: Gives a straight verdict on outreach numbers instead of saying it depends. Use when asked whether a reply rate, open rate, connection acceptance rate or positive reply rate is any good, why a campaign is getting no replies, how a campaign compares against published benchmarks, or to audit a set of campaign statistics. Use before offering any opinion on whether a figure is good or bad.

Read the whole file Download .md

Using them

What a skill is, and what to do with the file

If you have not used one before, this is the whole of it. There is no runtime, no dependency and nothing to keep updated.

Question	Answer
What is it, physically	One markdown file called SKILL.md, in a folder named after the skill.
What is in it	Frontmatter giving the name and the description that decides when it is used, then instructions in plain prose.
Where does it go	Wherever your assistant reads skills from. For Claude Code that is a skills folder in your home directory or your project.
How is it triggered	By the description. The model reads it, matches it against what you asked for, and follows the file when the two line up.
What does it cost	Nothing, and there is nothing to keep paying for. It is a text file.
Can I edit it	Yes. It is prose, not code, and the parts that are judgement calls say so.

There is no one-line install command for these yet, because that needs a public repository we have not set up. The zip does the same job: unpack it and copy the folders you want.

Two things we will not do

So you know what you are taking

No measured numbers of ours are inside a skill. A blog post can carry a figure and be corrected later. A skill gets copied onto someone else's machine, where nothing can be corrected. So the worked examples are illustrative and say so in the line above them. Where a benchmark table appears it is the upstream author's published figure, labelled as theirs, with a note that we have not measured it. Our own measured figures are published, dated, in the case study, where they can be updated.

No sending tool is named in any file. Each one asks for your own provider's limits or signals rather than assuming a product, and the ones that arrived naming a product had that taken out. The only vendor name left anywhere in a skill is in its Credits, where the original author is, because removing that is the one edit we are not entitled to make.

Questions about the library

Asked before anybody downloads anything.

It is a markdown file describing a task well enough that a model can carry it out consistently: what to ask for, how to work it out, what to report, and what not to claim. You put it where your assistant can read it, and it is used when the work matches its description. There is nothing to run and nothing to install beyond the file itself.

These come out of running outbound for clients, which is described on the cold email page, and priced on the pricing page. Neither is a condition of using anything here.

Claude is a product of Anthropic. TheCold.Ai is not affiliated with, endorsed by or connected to Anthropic. The adapted skills are used under the MIT licence and their original authors are credited on each page; those authors do not endorse this library.

11 skillsfor the jobs outbound is made of.

Everything in one zip

11 skills, grouped by the job they do

Targeting

Offer and angle

Writing and replies

Measuring, capacity and compliance

What a skill is, and what to do with the file

So you know what you are taking

Questions about the library

What does a Claude skill actually do?

Are these really free, and what is the catch?

Did you write all of these?

Why these ones and not the whole of the source library?

Do I have to give you an email address?

Can I change them, or use them with clients?

Are you affiliated with Anthropic?

Found the cap lower than you expected?

11 skills
for the jobs outbound is made of.