MCP App Store
by KeenEthics
Pageree icon

Pageree MCP Server for ChatGPT

By Pageree

Overview

Pageree lets your assistant take a landing page all the way to a business result: live on a real URL, collecting leads, and improving from how visitors actually behave. Describe the offer and the assistant builds the page, shows you a private preview, and publishes it when you approve. Add a form of any shape; Pageree validates submissions, blocks spam, keeps every lead, and delivers each one to your email, a webhook, or your CRM. Connect your own domain and the certificate is handled for you. Once visitors arrive, ask how the page is doing. The assistant reads traffic, sources, and section-by-section behaviour, such as where people stop scrolling and which calls to action get clicked, and can make a focused change, republish, and compare results. Every publish is a version, so any change can be rolled back. Built-in skills teach the assistant how to structure a page that converts, how to write the copy, and how to read the analytics, so it produces good pages without you managing the details. Nothing is public until you say so, and every page is checked against Pageree's content policy before it goes live.

How to connect

  1. ChatGPT: Open the Pageree listing in the ChatGPT app directory and choose Connect. ChatGPT listing Pageree exposes 39 tools here.

Tools

audit_page

ChatGPT
Audit a LIVE published page and get a scored, prioritized report across conversion, performance, SEO, mobile and accessibility — the "fully optimized" checkpoint. Run it while improving a page (before it has traffic); pair with get_analytics once it does. Returns markdown.

custom_domain_remove

ChatGPT
Disconnect a page's custom domain. The hostname stops serving the page immediately; the subdomain URL keeps working. Ask the user before calling it. Needed before the same hostname can be connected to another page.

custom_domain_set

ChatGPT
Connect a custom domain to a published page: a hostname the USER typed (never invent one — it must be a domain they own). Returns the CNAME record they must create at their DNS provider; SSL certificates are issued automatically after that. Then poll custom_domain_status once the user says the record is in. The subdomain URL keeps working alongside the custom domain.

custom_domain_status

ChatGPT
Check a page's custom domain: which hostname is connected, whether DNS verification and the SSL certificate are done, and what Cloudflare sees if not. Poll it after the user says the DNS record is in; "active" means https://<hostname> is live.

discard_draft

ChatGPT
Throw away the page's unpublished draft and stop its preview URL resolving. The live page is NOT affected — if the page is published it keeps serving its current version. Use this when the user abandons a direction and wants to start over, not to "clean up" after publishing: publishing already consumes the draft. Ask the user first. The draft is the only copy of that work and it cannot be recovered.

generate_image

ChatGPT
Generate a custom image for the page and return a hosted WebP URL to drop straight into an <img> src. Pass the EXACT pixel size the layout needs — the image is rendered and resized to it, so the page ships no oversized bytes. Use for hero art, backgrounds, illustrations and abstract visuals. Prefer the user's own imagery when they have it (logos, product shots, team photos convert better); prefer CSS for plain gradients or geometric shapes, which need no image at all.

get_analytics

ChatGPT
Read real analytics for a published page: traffic over the last 30 days (visits per day, sources, countries, browsers, devices), then engagement, scroll depth, bounce rate and per-section click/exit/time distribution — the data to drive the optimize loop (find the weakest section, change it, republish, compare). Optionally filter the behavioral part by device. Returns markdown.

get_draft_html

ChatGPT
Read back the full HTML of a page's unpublished draft — the work in progress behind its preview URL. Use it before editing a draft you did not write in this conversation: save_draft takes the COMPLETE document and replaces the draft with it, so without reading it first you would rewrite the page from scratch and lose whatever is already there. For the version visitors are actually seeing, use get_published_html.

get_page_information

ChatGPT
Look up one page and what state it is in: whether it is published and at which version, and whether it has an unpublished draft with a preview URL. CALL THIS BEFORE EDITING a page you did not just create — if a draft is already open, saving replaces it, so you need to know whether it holds work the user still wants. Accepts anything that identifies the page: its pageId, its slug, its Pageree URL, a connected custom domain, or a draft preview URL.

get_previous_version_html

ChatGPT
Read back the full HTML of an EARLIER published version of a page. Versions are immutable, so this is what that version actually shipped. Two uses: recovering something an later edit removed without rolling the whole page back, and the optimize loop — read the version that converted better, compare it against the live one from get_published_html, and see what the change actually was. Use list_versions to find the label. To make an old version live again, that is rollback, not this.

get_published_html

ChatGPT
Read back the full HTML of the version a page is LIVE with right now. This is the starting point for editing any page you did not build in this conversation: save_draft takes the COMPLETE document, so to change one heading you need the other 900 lines to send back with it. Read this, make the edit, then save_draft the whole thing. Returns the stored source, so it carries no analytics pixel — never add one. If the page has an open draft, that is newer than this; read it with get_draft_html instead.

get_skill_analytics

ChatGPT
How to READ get_analytics: what every behavioral metric means (sessions, click/bounce rate, scroll/reading behavior, per-section click/exit/time distribution) and how to spot the weak section. Load before interpreting analytics or planning an optimization.

get_skill_assets

ChatGPT
How to source page imagery: search_icons, generate_image and upload_asset, what size to ask for, when to use CSS or the user's own images instead, how to reuse what the account already has, and how to embed the results.

get_skill_best_practices

ChatGPT
How to build a landing page that CONVERTS: section composition and order, copywriting rules, and visual design principles. Load before designing or writing copy.

get_skill_content_policy

ChatGPT
What content can be published on Pageree hosting: prohibited and borderline categories (rejected at publish), and how to stay clearly on the allowed side.

get_skill_conversion_optimization

ChatGPT
The optimize loop: turn a weak analytics signal into ONE concrete page edit, republish, and measure the delta. Symptom→cause→fix playbook for low conversion, high bounce, scroll drop-off, weak CTA. Load after reading analytics, before editing to improve performance.

get_skill_custom_domains

ChatGPT
REQUIRED before using the custom_domain_* tools: how to connect a user's own domain to a published page — DNS instructions, SSL status, apex-domain caveats.

get_skill_delivery_integrations

ChatGPT
REQUIRED before changing where form submissions go (set_delivery): how to research the target service, build the field mapping, and connect it WITHOUT ever asking the user for credentials.

get_skill_discovery

ChatGPT
DO THIS FIRST for any new landing page: how to research the business, interview the user (only the gaps), and present a build plan the user approves before you build.

get_skill_form_management

ChatGPT
REQUIRED before adding a form to a page: the exact form markup contract (data-form, action, honeypot), how submissions are validated and delivered, and how to show success/error states.

get_skill_image_optimization

ChatGPT
For agents with a local filesystem: resize and convert the user's image files to webp with the provided script BEFORE uploading them with upload_asset. Chat assistants without files should skip this and pass the image to upload_asset directly.

get_skill_overview

ChatGPT
START HERE: what Pageree can do, the landing-page workflow, and which skill/tool to use at each step. Load this first when starting any landing-page task.

get_skill_page_structure

ChatGPT
REQUIRED before writing page HTML: the exact markup rules a Pageree landing page must follow (file layout, section markup for analytics, technical requirements).

get_skill_publish

ChatGPT
REQUIRED before saving or publishing a page: the draft → preview → publish workflow, how to store pageId/slug locally so later saves update the same page, changelogs, versions and rollback.

get_subprocessors

ChatGPT
Returns the current list of parties that process form submissions for a page — Pageree, our infrastructure sub-processors, and the destination this page delivers to. REQUIRED when writing a privacy policy: paste the entries verbatim and never recall, guess or invent them. Pass pageId to include that page’s configured delivery destination.

get_uploaded_page_assets

ChatGPT
List the images hosted for this account. With a pageId, the ones that page uses — an asset is attached automatically when you publish a page whose markup points at its URL. WITHOUT a pageId, the ones not on any page yet: images generated or uploaded but never published, which is where a wrong crop or an abandoned draft ends up. The two listings do not overlap. Use it before generating a new image — the user may already have one that fits — and to find the id remove_asset needs.

list_pages

ChatGPT
List the account's landing pages, most recently touched first: pageId, slug, live URL, current version, and whether each has an unpublished draft. Use it when you do not know a pageId — a new conversation, a lost .pageree file, or a user who refers to a page by name. Then pass the pageId to get_page_information or save_draft. A page showing a draft has work in progress that nobody has published; worth telling the user about.

list_versions

ChatGPT
List all published versions of a page: version number, changelog, date, and which one is currently live. Use before rollback or to review publish history.

publish

ChatGPT
Make the page live at https://<slug>.pageree.app by publishing its draft. Save the page with save_draft first — this tool takes no HTML, it promotes whatever the draft currently holds. PUBLISHING IS THE USER'S DECISION: it is the moment the page becomes public. NEVER call it in the same turn as save_draft. Always show the preview URL first and wait for the user to look at it and approve in a later message — even if they opened with "get it live", "host it" or "publish it": that tells you the goal, not that they have seen this page. Each publish is an immutable version; use list_versions and rollback to manage them. The draft is consumed — continuing to edit means saving a new one.

remove_asset

ChatGPT
Permanently delete a hosted image — the file and its record. Use it to clear out images the user does not want kept: a wrong crop, an abandoned draft, a photo they asked you to remove. Get the id from get_uploaded_page_assets. An image a published page uses is refused, because deleting it would leave a broken image on the live page; republish that page without it first, or pass force. This cannot be undone.

rollback

ChatGPT
Make an earlier version of a page live again (versions are immutable — this just repoints the live URL). Use list_versions to find the version to roll back to.

save_draft

ChatGPT
Save a landing page as a draft and get back a private preview URL to show the user. NOTHING is public until you call publish: a draft is not indexed and records no analytics. Call get_skill_publish BEFORE first use (the workflow, and how to store pageId locally) and get_skill_page_structure BEFORE writing the HTML — saving validates against that spec and rejects the page if it does not comply. A page is ONE HTML document with its CSS and JS inline. Images must be hosted URLs from generate_image or upload_asset — never separate files, never base64 data URIs. The first save creates the page and returns its pageId: SAVE IT (see get_skill_publish) and pass it on every later save, or you will create a second page instead of updating this one. Saving again replaces the draft in place and keeps the same preview URL, so a link you already gave the user keeps working.

search_icons

ChatGPT
Search for SVG icons by name. Returns the inline SVG markup for the best match per query — embed it directly in the page HTML. Icons inherit color via currentColor and scale via CSS. Use for buttons, feature lists, navigation, contact rows, etc.

set_delivery

ChatGPT
Configure where a page's form submissions are delivered: type "email" (an address the user chose) or type "http" (any CRM API / Zapier / Make / n8n webhook / custom backend, with headers and a field mapping). Load get_skill_delivery_integrations BEFORE using this — it defines the required workflow: research the target API, build the mapping yourself, confirm details with the user, and NEVER ask for credentials in chat — create without secrets and send the user the returned settingsUrl to add auth securely.

set_form_rules

ChatGPT
Set the rules that mark a page's form submissions invalid instead of delivering them, such as repeat email addresses, countries, email domains, throwaway mailboxes, domains that cannot receive mail (every rule type is listed below). The visitor always sees success; an invalid submission is kept 30 days in the Submissions inbox, not delivered and not counted as a conversion, and the owner can release it. This REPLACES the whole list: call get_page_information first and send back every rule to keep, with its id. Load get_skill_form_management before using this. Rule types (at most 20 rules per page): - country — Marks a submission invalid by the visitor's country, from IP geolocation. "allow" accepts only the countries listed; "block" refuses the ones listed. A visitor whose country is unknown passes. VPNs and private relays move people, so a strict allow list will catch some real visitors: the owner can release them. config: {"mode":{"type":"string","enum":["allow","block"]},"countries":{"type":"array","items":{"type":"string"},"description":"ISO 3166-1 alpha-2 codes, e.g. [\"US\", \"CA\", \"GB\"]."}} - email_domain — Marks a submission invalid by the email address's domain. "allow" accepts only the domains listed; "block" refuses the ones listed. An entry covers the domain and its subdomains, so "acme.com" also matches mail.acme.com, and "edu" (or ".edu") matches every address under .edu. config: {"field":{"type":"string","description":"The form field with type=\"email\"."},"mode":{"type":"string","enum":["allow","block"]},"domains":{"type":"array","items":{"type":"string"},"description":"Domains, e.g. [\"acme.com\", \".edu\"]."}} - disposable_email — Marks a submission invalid when the email address is a throwaway mailbox (mailinator.com and about ten thousand others, from a public list). config: {"field":{"type":"string","description":"The form field with type=\"email\"."}} - mail_server — Marks a submission invalid when the email address's domain cannot receive mail (it does not exist or has no mail server), which catches made-up domains and typos like gmial.con. It does not check that the mailbox itself exists. config: {"field":{"type":"string","description":"The form field with type=\"email\"."}} - email_verification — Checks the email address with the owner's own verification provider (ZeroBounce, Kickbox, Emailable, Bouncer) and marks the submission invalid when the provider says the mailbox does not exist; with riskyIsInvalid, also catch-all domains, role and disposable addresses. Each check spends the owner’s credits with that provider. The owner adds the provider’s API key in the console, never in chat. List more than one provider to fall back when the first cannot answer; when none can, the submission is delivered and the owner is emailed. config: {"field":{"type":"string","description":"The form field with type=\"email\"."},"providers":{"type":"array","items":{"type":"string","enum":["zerobounce","kickbox","emailable","bouncer"]},"description":"In order of preference, at most 3, e.g. [\"zerobounce\"]."},"riskyIsInvalid":{"type":"boolean","description":"Also mark catch-all, role and disposable addresses invalid. Default false."}} - phone_verification — Checks the phone number with the owner's own provider (Twilio or Veriphone) and marks the submission invalid when it is not a real number, or when its line type is one the owner blocks (for example "voip"). A number without a country code is read as the visitor’s country’s. Each check spends the owner’s credits with that provider (Twilio charges for line types). The owner adds the provider’s credentials in the console, never in chat. config: {"field":{"type":"string","description":"The form field with type=\"tel\"."},"provider":{"type":"string","enum":["twilio","veriphone"]},"blockedLineTypes":{"type":"array","items":{"type":"string","enum":["mobile","landline","voip","toll_free","premium","other"]},"description":"Line types to mark invalid. Default []."}} - duplicate_email — Marks a submission invalid when the same email address was already accepted in the la…

set_goal

ChatGPT
Define the page's conversion goal so analytics can report a real conversion RATE (conversions ÷ sessions), not just behavioral proxies. The goal is either a CTA click (matched by button/link text, url, and/or section) or a form submission (by name, or any form). Changing the goal re-scores all past data instantly — nothing is migrated. Use action "clear" to remove the goal.

upload_asset

ChatGPT
Host an image the user already has — a logo, product shot, screenshot or team photo — and get back a URL to use in the page. Real imagery of the user's own business converts better than anything generated, so prefer this whenever they can supply a file. Choose ONE transport: sourceUrl if the image is already on a public https URL (we fetch it); no argument at all to get a one-time upload URL to PUT the file to with curl (best when you have a shell and a file on disk); or base64 for a small file you can inline. Whatever arrives is resized to fit, converted to WebP and stored — including SVG, which is rasterised. The asset is attached to a page automatically when you publish a page that uses its URL.

whoami

ChatGPT
Returns the authenticated Pageree user and the account their pages belong to. Use it to verify the connection, and to confirm which account you are working in before publishing.

Capabilities

Writes

Example Prompts

Click any prompt to copy it.

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/pageree"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/pageree)