Guide

Docs for AI and agents

Give ChatGPT, Claude, Cursor, Codex, Grok and other AI tools accurate Emailit context with llms.txt, Markdown pages, the MCP server and the Emailit plugin.

Updated Oct 2, 2026

AI assistants give better answers about Emailit when they read the current docs instead of guessing. Emailit gives them three layers of help:

Layer	What it gives the assistant	Start here
Docs as Markdown	Every docs page and the main pages of emailit.com as plain Markdown, plus llms.txt indexes.	This page
MCP server	109 tools that act on your workspaces: send email, set up domains, run campaigns and more.	MCP server
Plugin and skills	Emailit know-how for Claude Code, Codex, Cursor and Grok Build: setup flows, deliverability checks, SDK patterns and safe handling of keys.	Plugins and skills

Use the docs alone to answer questions and write code. Add the MCP server when the assistant should also do the work in your account.

Files for AI tools

File	What it contains	Use it for
`https://emailit.com/llms.txt`	A short summary of Emailit: products, pricing, credits per action, the MCP server and key links.	Giving an assistant a quick overview of what Emailit is.
`https://emailit.com/llms-full.txt`	The main pages of emailit.com in one file: products, pricing, credits, programs, answers and comparisons.	Questions about plans, products, partners, referrals and how Emailit compares.
`https://emailit.com/docs/llms.txt`	An index of every docs page, grouped by section, with a one-line summary and a link to its Markdown version. It also lists key facts such as the API base URL, SMTP settings and the MCP server URL.	Letting an assistant find and fetch the pages it needs.
`https://emailit.com/docs/llms-full.txt`	The full text of every docs page in one file.	Loading all of the docs into a project, a knowledge base or an editor’s docs index.

These files follow the llms.txt convention and are rebuilt every time the site or the docs change.

Markdown version of any page

Add .md to a page’s URL, without the trailing slash, to get it as Markdown. This works for the docs and for the main pages of emailit.com:

Page	Markdown
`https://emailit.com/docs/quickstart/api/`	`https://emailit.com/docs/quickstart/api.md`
`https://emailit.com/docs/webhooks/request-signature/`	`https://emailit.com/docs/webhooks/request-signature.md`
`https://emailit.com/docs/mcp/tools/`	`https://emailit.com/docs/mcp/tools.md`
`https://emailit.com/pricing/`	`https://emailit.com/pricing.md`
`https://emailit.com/products/email-api/`	`https://emailit.com/products/email-api.md`
`https://emailit.com/vs/sendgrid/`	`https://emailit.com/vs/sendgrid.md`
`https://emailit.com/`	`https://emailit.com/index.md`

Each HTML page points to its Markdown version with <link rel="alternate" type="text/markdown">, so agents and crawlers can find it without guessing the URL.

Interactive parts of a page, such as tabs, callouts and plan tables, become plain Markdown, so code samples for every language are included. For the API reference, each resource also has one combined file with all of its endpoints, for example https://emailit.com/docs/api-reference/emails.md for the Emails API.

Page actions

Every docs page has a Copy page button next to the title. Select it to copy the page as Markdown, ready to paste into a chat. The arrow next to it opens more options:

Action	What it does
Copy as Markdown	Copies the page’s Markdown to your clipboard.
View as Markdown	Opens the `.md` version in a new tab.
Open in ChatGPT	Starts a ChatGPT conversation that asks it to read the page’s Markdown, so you can ask questions about it.
Open in Claude	Starts a Claude conversation with the same prompt.

Use the docs in your tools

Chat assistants. Paste a page with Copy page, or ask the assistant to read a .md URL. For broad questions, point it at https://emailit.com/docs/llms.txt and let it pick pages.
Projects and knowledge bases. Add https://emailit.com/docs/llms-full.txt as a source in a Claude project, a custom GPT or your team’s internal assistant. Refresh it from time to time to pick up changes.
Code editors. Add https://emailit.com/docs/llms-full.txt (or https://emailit.com/docs/) as a documentation source in Cursor or a similar editor, then reference it when you ask for Emailit code.
Coding agents. Install the Emailit plugin in Claude Code, Codex, Cursor or Grok Build. Its skills load Emailit setup and SDK guidance only when a task needs it, and point the agent at these Markdown pages for details.

Prefer a single page’s .md file for focused questions: it’s smaller than the full file and keeps the assistant on topic.

Combine the docs with the MCP server

The docs tell an assistant how Emailit works; the MCP server lets it act on your workspace. Used together, an assistant can read the rules and then do the work. For example, in Cursor, Codex or Claude Code with the Emailit MCP server connected:

Text

Read https://emailit.com/docs/webhooks/request-signature.md and
https://emailit.com/docs/webhooks/event-types.md.

Then create an Emailit webhook named "Bounces" that sends email.bounced
and email.complained events to https://acme.com/webhooks/emailit, and
add an Express route at /webhooks/emailit that verifies the signature
with the webhook's secret and marks bounced addresses in our users table.

The assistant reads the signing rules from the docs, calls create-webhook to set up the endpoint and get the signing secret, and writes code that matches.

Tips

Check the date. Each page shows when it was last updated. If an assistant’s answer conflicts with a page, trust the page.
Keep secrets out of prompts. The docs never need your API key. Prefer OAuth for the MCP server, and keep keys in your environment, not in chat messages or committed files.
Report gaps. If an assistant can’t find something in the docs, it’s probably missing for people too. Tell us at support@emailit.com.

For crawlers and tool builders

robots.txt allows search engines and AI crawlers, including GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-User, PerplexityBot and Google-Extended. Only the dashboard API and admin paths are excluded.
Markdown responses use text/markdown, and the llms.txt files use text/plain. All of them are static files, cached at the edge.
Pages carry schema.org structured data (TechArticle for docs, plus breadcrumbs), a canonical URL and an updated date.
The MCP server describes every tool with a JSON schema, a title and annotations such as read-only and destructive, and returns structured results. See the tool reference.