---
title: "Cold Email Playbook Claude Skill: Diagnose Any Drop"
description: "A Claude skill that works an open-ended cold email problem the way our operators do: checks in order, findings weighed by cost, and one outcome with the…"
url: https://warminboxes.com/skills/cold-email-playbook
breadcrumb: "Home > Claude Skills > Cold Email Playbook"
---
Claude skill

# Cold Email Playbook

Why did my replies drop, and what do I do?

The other skills each run one job. This one decides which to run, in what order, and what a finding is worth, from the WarmInboxes playbook, the operator notes and the published studies. Authentication first, then bounces, then replies against your own baseline, and new infrastructure only when the evidence says degraded or out of capacity.

## What it runs

- **The symptom** — Matched to a playbook scenario: reply drop, blocklist, Outlook junk, Gmail spam, bounces, burned domain, setup or scale

- **Authentication** — SPF, DKIM, DMARC and MX, before anything else

- **Bounces** — A bad list cleaned; a sender problem rested, not slowed

- **Replies against your baseline** — Asked for, because no server can read your campaign history

- **Evidence** — The operator notes and the studies, each labelled with what kind of evidence it is

## What you get back

- One outcome: healthy, fixable, unclear, degraded or capacity

- The evidence behind it, each piece labelled as dataset, operator rule, observation or hypothesis

- Each note’s “do not” list, passed on as plainly as its steps

- The single next action, with study figures quoted from the study rather than from memory

## Install it

Best installed with the plugin, which also connects WarmInboxes Free MCP, the tools this skill runs. On its own it is one file that falls back to the free WarmInboxes API, so there is no key and nothing to sign up for.

 /plugin marketplace add Huz-ai-fa/warminboxes-claude-skills
/plugin install warminboxes@warminboxes

Or this one skill on its own:

 mkdir -p ~/.claude/skills/cold-email-playbook
curl -sL https://raw.githubusercontent.com/Huz-ai-fa/warminboxes-claude-skills/main/cold-email-playbook/SKILL.md -o ~/.claude/skills/cold-email-playbook/SKILL.md

[View it on GitHub →](https://github.com/Huz-ai-fa/warminboxes-claude-skills/tree/main/cold-email-playbook)

## Then just ask

You do not name the skill. Its description tells Claude when it applies.

- "Our reply rate halved this month, what should I check?"

- "We are listed on a blocklist, does it matter?"

- "What does your research say about Microsoft tenants for cold email?"

Every answer cites the check behind it, so anything Claude tells you can be re-run by hand. A clean result is reported clean, with nothing to buy.

[All Claude skills →](https://warminboxes.com/skills)

## Do it by hand

The same checks as free tools in a browser.

- [Deliverability Checker](https://warminboxes.com/deliverability-checker)

- [Bounce Analyzer](https://warminboxes.com/bounce-analyzer)

- [Blacklist Checker](https://warminboxes.com/blacklist-checker)

## The other skills

One page each.

- [Inbox Checker](https://warminboxes.com/skills/inbox-checker) — Is this domain burned?

- [Domain Audit](https://warminboxes.com/skills/domain-audit) — Is this domain set up correctly to send from?

- [List Segmentation](https://warminboxes.com/skills/list-segmentation) — Who hosts the people on this list?

- [Infrastructure Planner](https://warminboxes.com/skills/infrastructure-planner) — How many mailboxes do I need, and what does it cost?

- [Bounce Diagnosis](https://warminboxes.com/skills/bounce-diagnosis) — What do these bounces mean, and is it the list or me?

- [Cold Email Copy](https://warminboxes.com/skills/cold-email-copy) — Is this email going to land in spam?

- [Lead Lists](https://warminboxes.com/skills/lead-lists) — Is this list ready to send to?

- [Warm Rotation](https://warminboxes.com/skills/warm-rotation) — How do I make my inboxes last longer?

- [Infrastructure API](https://warminboxes.com/skills/infrastructure-api) — How do I connect my WarmInboxes account to Claude?