MCP documentation
Email Love is a native MCP server. Point any AI tool at one URL and it can search the library, pull cleaned HTML, walk lifecycle journeys, and read your own follows, likes and collections. Prefer a plain REST API? See the API docs.
Overview
MCP (Model Context Protocol) is an open standard that lets AI assistants connect to outside tools and data. The Email Love MCP brings the entire library, 570,000+ marketing emails, lifecycle journeys, and brand-level analytics, directly into Claude, ChatGPT, Cursor, Claude Code, and any MCP-enabled assistant. In plain terms: it turns your AI into an email marketing expert that has actually seen thousands of real campaigns, and can pull up the exact ones you need, in the middle of whatever you are working on. No tab-switching, no copy-paste.
Free with an Email Love account. Register free →
Plus the Email Love skills
The MCP gives your agent the library; skills teach it what to do with it. They’re free and open source (MIT), and install alongside the MCP in one pass: see Set up your agent.
| Pack | What it does | Source |
|---|---|---|
| Lifecycle calendar 1 skill | Builds a stack-ranked year of campaign moments in Notion, with what competitors actually sent at each moment and how deep they discounted. Also needs the Notion connector. | lifecycle-calendar |
| ESP skills 10 skills | Write, review, and debug personalization for the platform you send from: Braze Liquid, Customer.io Liquid, HubSpot HubL, Iterable Handlebars, Klaviyo, Marketo Velocity, MoEngage Jinja, Sailthru Zephyr, SFMC AMPscript, and Zeta ZML. | esp-skills |
| Figma skills 5 skills, 4 in Codex | Build, repair, audit, and migrate export-ready emails in Figma from your Email Love design system. Also needs the Figma MCP and the Email Love Figma plugin. | claude-skills codex-agents |
Set up your agent
Connect the library and install the Email Love skills in one pass. It’s a one-time setup and the URL is always https://chat.emaillove.com/mcp: you sign in with your Email Love account and click Allow, so there’s no secret key to paste. Pick your agent and run the commands, or copy the setup brief and let your agent do it.
Sign in once in the browser with a free Email Love account. No keys to paste. Skills are open source (MIT).
Works with any MCP client that supports HTTP with OAuth, including Figma’s AI agent through its custom MCP setting. In Claude.ai and Claude Desktop, results render as inline cards and carousels with real email screenshots; CLI tools like Claude Code return the structured data.
Tools
You don't memorize commands, you just ask. Under the hood, these tools do the work:
Every tool is free for now, with no monthly limit, at up to 600 requests a minute per account. All of them are read-only except request_brand, which only adds a brand to our collection queue.
| Tool | Arguments | What it does |
|---|---|---|
| search_emails | query?, brand?, category?, since? (date, e.g. 2026-08-01), page? (default 1), limit? (1-25, default 10) | Search the email library. Returns matching email summaries with next_page. With a brand and no query, results come newest first, so brand + since + page walks a brand's history. |
| fetch_email | id (integer) | Full details for one email: brand, categories, lifecycle stages, screenshot. |
| get_email_html | id (integer) | Full sanitized HTML source of one email: study real structure, responsive styles, and footer markup, or use it as the reference for ESP templating work. |
| search_brands | query?, limit? (1-25, default 10) | Find brands. Omit query for the top brands by email count. |
| list_journeys | limit? (1-50, default 30) | List brands that have a curated lifecycle journey. |
| get_journey | brand (slug), emails_per_stage? (1-10, default 5) | One brand's lifecycle flow, by stage. The most differentiated tool. |
| get_brand_insights | brand (slug) | Brand-level analytics: ESP, send timing, subject-line patterns, content metrics, deliverability. |
| search ยท fetch | query / id | Convenience aliases for search_emails / fetch_email. |
Your account
These read the brands you follow, the emails you have liked, and your collections on emaillove.com. They only ever return the account you signed in with.
| Tool | Arguments | What it does |
|---|---|---|
| list_followed_brands | limit? (1-50, default 50) | The brands you follow on Email Love, most recent follow first, with logos and email counts. |
| list_followed_brand_emails | since? (date, e.g. 2026-10-01), limit? (1-25, default 12) | The newest emails from every brand you follow, newest first. Pass since to see only what is new. |
| list_liked_emails | limit? (1-100, default 50) | The emails you have liked on Email Love. |
| list_collections | none | Your saved collections: names, descriptions and how many emails each holds. |
| get_collection | collection (id or name), limit? (1-100, default 50) | One of your collections and its emails. |
Follow a brand with the + Follow button on its page or on any email card on emaillove.com. You can follow up to 50 brands.
Brand requests
Missing a brand? Ask for it. Our signup agent subscribes to requested brands twice a day, and their emails appear in the library as they send them, usually within a few days. Up to 50 requests per account per day, shared with the REST API. Each brand files into one of the library’s categories (Apparel, Automotive, Baby, Beauty, Beverage, Cannabis, Crypto, Digital Products, Electronics, Food, Health, Higher Education, Home, Jewelry & Watches, Luggage, Non-Profit, Pets, Restaurants, SaaS, Sports Teams, Travel, Video Games); leave category out and we pick the closest one. Email Love does not collect adult content, so those requests are refused.
| Tool | Arguments | What it does |
|---|---|---|
| request_brand | brand, website, category?, signup_url?, notes? | Ask Email Love to start collecting a brand. A brand already in the library comes straight back; otherwise the request is queued and our signup agent subscribes. |
| list_brand_requests | limit? (1-200, default 50) | Your requests, newest first, with each one's status and how many requests you have left today. |
Statuses: queued, signed_up (subscribed, waiting for the first email), first_email_captured (in the library, with its brand page) and could_not_subscribe (for example, the brand only offers signup through an app or an account).
Example prompts
A few that show what it can do:
- “Compare REI and Patagonia’s email programs: how often they send, what ESP they use, and how their subject lines differ.”
- “Show me 3 post-purchase emails from DTC apparel brands. Which one has the cleanest design?”
- “Walk me through MeUndies’ winback flow.”
- “Pull the HTML of Patagonia’s latest email and show me how they build their footer.”
- “What words and emojis does Fabletics use most in its subject lines, and how long are they on average?”
- “When do Lululemon, Alo Yoga and Vuori send their emails? Compare their days and times.”
- “What have the brands I follow sent since Monday?”
- “Open my Welcome ideas collection and tell me what those emails have in common.”
Privacy
When you connect, Email Love receives only what it needs: your Email Love user id and email (so we know whose account is asking) and your connecting IP for a couple of minutes (for rate-limiting, then discarded). We do not receive your conversation, your prompts, or anything beyond the specific tool call your assistant makes. The account tools read your own follows, likes and collections and nothing from any other account. A brand request stores the brand and your account id so you can check on it. Your access token lasts 30 days; revoke the connector in your AI tool any time, or change your Email Love password to invalidate all sessions.
Questions? hello@emaillove.com