Skip to content

MCP connector

Ask your AI about your whole life.

Framezone speaks MCP, the open standard AI assistants use to talk to other apps. Connect Claude, ChatGPT, Cursor, VS Code or any other MCP client, and it can read your tasks, habits, goals, money, reading and training, and add or update things when you ask.

Start free Set it up
  • 62 tools
  • You approve access
  • Free during early access

Example conversation

How did my week go?

You finished 14 tasks, including two of your three Top 3 items on Friday. Meditation is on a 47 day streak, but you skipped reading four days out of seven. Spending came in at 312 euros, mostly groceries and one train ticket.

Move the physio booking to Monday and make it high priority.

Done. "Book the physio" is now due Monday with high priority. It will show up in Monday's list.

What you can ask

Plain questions, real answers.

Talk to your assistant the way you'd talk to a friend who can see your dashboard. A few things people ask, by area.

Tasks
  • "What should I focus on today?"
  • "Add a task to call the dentist tomorrow, high priority."
Habits
  • "Log my meditation for today."
  • "Which habits did I miss this week?"
Money
  • "What are my active subscriptions and what do they cost a month?"
  • "Log a transaction: 4.50 for coffee."
Fitness
  • "Log a 5 km run for today."
  • "How is my weight trending against my goal?"
Reading
  • "Which books am I reading right now?"
  • "Save this quote from the book I just finished."
Journal
  • "Write my evening journal: productive day, energy 4 out of 5."
  • "How was my mood over the last two weeks?"

What your assistant can do

62 tools across every area.

Each tool is one thing your assistant can read or change in your account. It picks the right one from what you ask, so you never need to know their names.

Tasks5

Create, update, complete and list

Projects7

Projects plus site analytics

Habits4

Create, log and check streaks

Goals3

Create, update and track progress

Focus timer2

Log sessions, see your focus time

Journal2

Read and write daily entries

Notes3

Create, list and delete

Books & quotes6

Your shelf and saved quotes

Fitness9

Workouts, sets, body and food

Finance5

Accounts, transactions, subscriptions

Portfolio2

Holdings and trades

Contacts3

Create, update and list people

Bookmarks2

Save and remove links

Wishlist3

Add, update and list items

Achievements1

See what you have unlocked

Memory3

Things your assistant should remember about you

Search & summary2

Search everything, weekly summary

Set it up

Connect the assistant you already use.

You need a Framezone account and an assistant that supports MCP. It takes about two minutes and costs nothing extra. Add a few things to Framezone first, so there is something to ask about.

Claude · Desktop app or claude.ai

  1. 1

    Open Settings, then Connectors

    In the Claude app or on claude.ai.

  2. 2

    Add a custom connector

    Name it Framezone and paste the URL below.

  3. 3

    Sign in and approve

    Claude opens Framezone. Log in, check the permissions and approve.

https://framezone.app/mcp

ChatGPT · Developer mode (beta)

  1. 1

    Turn on developer mode

    In ChatGPT settings. It is a beta feature, available on some plans.

  2. 2

    Add a new connector

    Choose a custom MCP server and paste the URL below.

  3. 3

    Sign in and approve

    ChatGPT opens Framezone. Log in and approve, then enable it in a chat.

https://framezone.app/mcp

Cursor · mcp.json

  1. 1

    Open your MCP settings

    Add this to ~/.cursor/mcp.json (or the project's .cursor/mcp.json).

  2. 2

    Sign in on first use

    Cursor opens your browser so you can log in to Framezone and approve.

{
  "mcpServers": {
    "framezone": { "url": "https://framezone.app/mcp" }
  }
}

VS Code · Copilot agent mode

  1. 1

    Add the server

    Put this in .vscode/mcp.json, or run "MCP: Add Server" and choose HTTP.

  2. 2

    Sign in on first use

    VS Code opens your browser so you can log in to Framezone and approve.

{
  "servers": {
    "framezone": { "type": "http", "url": "https://framezone.app/mcp" }
  }
}

Claude Code · Terminal

  1. 1

    Run this once

    The first time you use it, your browser opens so you can log in and approve.

claude mcp add --transport http --scope user framezone https://framezone.app/mcp

Any MCP client · Remote server

  1. 1

    Server URL

    https://framezone.app/mcp

  2. 2

    Transport

    Streamable HTTP

  3. 3

    Sign-in

    OAuth 2.1 with dynamic client registration. Scopes: read, write, delete.

https://framezone.app/mcp

Framezone follows the MCP standard, so any client that supports remote servers with OAuth sign-in should work. I test it end to end with Claude; if another assistant gives you trouble, tell me and I'll look into it.

Permissions & privacy

You decide what your AI can touch.

Your assistant never sees your password. It gets a token you approve, limited to what you allowed, and you can take it back.

Sign in with OAuth

You log in on Framezone itself. The assistant receives a token, never your password.

You choose the access

Read, write and delete are separate permissions. Untick write or delete on the consent screen and the assistant can only read.

Disconnect anytime

Remove the connector in your assistant whenever you like and it stops working.

Rate limited

60 requests a minute and 1,000 an hour per account, so nothing runs away.

"I never sell your data. When you use an AI assistant, only what you ask about is sent to it."
Sedat, maker of Framezone · How Framezone handles your data

Try it with your own dashboard.

Create an account, add a few things, then ask your assistant about them. Free during early access, MCP connector included.

Start free Explore features

Free during early access. No card needed.