Legal
Privacy Policy
Last updated: October 31, 2026
MCP Tutor is an AI-literacy training tool. Teaching, conversation, and assessment happen inside the AI assistant you connect it to (ChatGPT, Claude, Gemini, Grok or Perplexity); MCP Tutor is the progress-tracking and content-delivery layer behind it. This policy explains what data we collect, how we use and store it, who we share it with, and how long we keep it.
What we collect
- Account information. Your name, email address, and the organization you belong to, plus your role (member, manager, or admin) and plan.
- Training progress. Which chapters you have started or completed, timestamps, and scores.
- Demonstration records. The text you submit for a chapter’s demonstration and the written assessment of it. This is the core function of the product — recording your practice attempts.
- Playbook records. The playbooks you create (their titles, descriptions, and step content) and a log of your playbook runs — which playbook you ran, how many steps you completed, optional verification notes, and timestamps — used to power the usage view for your organization’s admins.
- Operational logs. Request metadata (timestamps, the tool invoked, latency, and error class) used to operate and debug the service. We do not log the content of your AI conversations.
- Website analytics. When you visit the MCP Tutor website, Vercel Web Analytics records each page view: the page’s address, the site that linked you here, the time, your approximate location (country, region and city), and your browser, operating system and device type. Before an address is sent we remove record IDs and everything after the “?” except campaign tags such as utm_source. Vercel tells visitors apart with a hash of the request rather than cookies and discards it after 24 hours; we see only aggregate counts.
- Signup source. When you create an account we record which signup form you used and, if you arrived from a link or ad, its campaign tags (such as utm_source or an ad click ID), the site that linked you here and the page you landed on. Your browser keeps these in local storage (not a cookie) until you sign up, and we store them with your account so we know which campaigns bring people to MCP Tutor.
We do not request or access your AI assistant’s memory, chat history, files, or any data outside the training records described above. We do not collect health data, and we do not sell personal data.
How we use it
- To deliver curriculum content and track your progress through it.
- To walk you through your playbooks and log runs for your organization’s usage view.
- To show managers and admins aggregate progress for their own organization.
- To authenticate you and operate, secure, and debug the service.
How we store and secure it
Data is stored in a Postgres database (Supabase). The server limits each account to its own records and each organization to its own members, and row-level security in the database adds a second layer for direct access from the browser. Connections use HTTPS/TLS. Access to the training server is authenticated with OAuth 2.0 — your AI assistant obtains a short-lived access token after you sign in and authorize the connection; tokens are never placed in URLs. Command-line tools that can only send a fixed header use a permanent connector token instead, which works only on the Gemini address.
Who we share it with (sub-processors)
We use the following service providers to run MCP Tutor:
- Supabase — database, authentication, and storage.
- Railway — hosting for the MCP server.
- Vercel — hosting for the website and web dashboard, and page-view analytics (Vercel Web Analytics).
- Resend — transactional email (sign-in codes).
- Stripe — payment processing for the Business plan; we never see your card number.
- Your AI assistant — Anthropic (Claude), OpenAI (ChatGPT), Google (Gemini), xAI (Grok) or Perplexity — the client you connect MCP Tutor to. Your conversation happens inside that assistant under its provider’s own terms and privacy policy; MCP Tutor receives only what the assistant sends to its tools (your demonstration response and the assistant’s assessment), never the rest of the chat.
We share data with these providers only as needed to operate the service. We do not sell or rent personal data, and we do not share it for advertising.
Retention
We keep account, progress, demonstration, and playbook records for as long as your account is active. When an account or organization is deleted, its progress, demonstration, and playbook records are deleted with it. Operational logs are retained for a limited period for debugging and security and then rotated out.
Your choices
You may request access to, correction of, or deletion of your personal data by contacting us.
Contact
Questions about this policy or your data? Email privacy@mcptutor.com (or support@mcptutor.com for general support).
Changes
We may update this policy as the service evolves. We will revise the “last updated” date above when we do.