Email Marketing Bible

The Email Marketing Bible turns an agent into an email marketing operator. It is split in two parts. Part A is the operating manual for the moments when the agent acts on a real account. Part B is a dense reference covering metrics, flow recipes, copywriting, segmentation, analytics, deliverability, testing, compliance, cold email, WhatsApp, SMS and RCS, platform selection, email design, 19 industry playbooks and mid-2026 benchmarks.

The working loop is read state, reason, act, verify. The agent reads lists, flows, recent campaigns, deliverability and suppressions first, changes one thing at a time and checks the result against real counts. Before any send it shows a pre-send packet (preview URL, audience size, suppressions, subject, preview text, send time, sender, unsubscribe, compliance risk) and waits for an explicit "send it".

The skill also carries an anti-slop copy protocol: one defensible opinion per email, varied sentence length, a lint list of overused words, and a rule that specific numbers and names beat generic phrasing. For design it covers a Discover, Define, Deliver process, safe email markup (MJML, React Email or Maizzle), a single-column layout up to 600px, dark-mode-safe colors and explicit text and button colors.

Use it when you are:

• building or editing automation flows such as welcome, abandoned cart, browse abandonment, post-purchase, win-back and sunset
• having an agent drive an ESP through MCP, a connector or an API (Klaviyo, Mailchimp, Resend, Nitrosend and similar) to create segments, flows or campaigns
• diagnosing deliverability: spam or Promotions placement, bounce or complaint spikes, SPF, DKIM and DMARC problems, warm-up and ESP migrations
• writing, reviewing or de-slopping copy, subject lines and calls to action
• critiquing a rendered email, or choosing an ESP, pulling benchmarks, checking CAN-SPAM, GDPR, CASL or Australian Spam Act rules, or planning cold email, WhatsApp, SMS or RCS

Do not treat it as a sending tool. It is guidance; actions run through the ESP's own MCP server, connector or API under that tool's permissions. It is not legal advice, and it does not replace a test send in a real inbox or human ownership of brand voice, strategy and the final send. Benchmarks, penalties, platform features and WhatsApp pricing are mid-2026 figures, so verify volatile facts first.

A task router maps each intent to a section and lists what to gather first. Building a flow needs the model, trigger, audience, offer and exclusions. Sending a campaign needs the segment, consent basis, copy, sender and timing. Deliverability work needs the domain, ESP, bounce and complaint rates and recent changes. Copy work needs audience, offer, voice and one real proof point.

Outputs are things like a prioritized gap list from an account audit, a flow scaffold, a tightened draft with changes listed, a deliverability diagnosis with a 2-4 week monitoring plan, or a staged draft with a verified segment count and the completed checklist. One example prompt from the source:

Audit my Klaviyo account for a DTC skincare brand doing $2M a year. What am I missing?

The expected result is a read-only pass over lists, flows and recent campaigns, then a gap list against the flow build order and the metric thresholds, with no writes.

The skill is platform-neutral and names the ESPs it compares: Klaviyo, Mailchimp, Customer.io, ActiveCampaign, HubSpot, Kit, Brevo, beehiiv, Omnisend, Resend, Iterable, Postup, Bento and Nitrosend, noting which offer an MCP server or connector. The author discloses that the guide shares a founder with Nitrosend. Field notes from running Nitrosend's sending, such as optimistic-concurrency versions on writes and re-asserting the brand before each write batch, are flagged as ESP-specific mechanics to map to your own platform. The only shell command in the skill is a local random seed string; it makes no network calls. Keep ESP API keys out of prompts, templates and logs.

Lifecycle and email marketers who want an agent to work inside their ESP without risking a live audience, ecommerce and SaaS teams building their first flows, developers wiring an ESP into an agent, and anyone triaging a deliverability drop. The hard gates make it suited to accounts where one wrong send has real cost: no send to more than one recipient without human approval, and a blocked send when authentication, unsubscribe, physical address or consent is missing, or complaint rate is at or above 0.1%.