MCP · Model Context Protocol

Your money, in every AI you use.

Connect your SpendixAI workspace to the assistant you already work in. Ask about spending, add expenses and upload statements from a conversation, with the access you choose.

Read only to startSign in with your SpendixAI accountDisconnect any time

One page for every AI agent.

Each page has the exact steps for that assistant, what you need on your side, and what to ask first.

Chat assistant

Claude

Add SpendixAI as a custom connector in Claude on the web or desktop and ask about your spending in conversation.

Set up Claude
Coding agent

Claude Code

One command adds SpendixAI to Claude Code, so you can ask about spending from your terminal.

Set up Claude Code
Chat assistant

ChatGPT

Add SpendixAI to ChatGPT as a custom MCP plugin and ask about your records in a chat.

Set up ChatGPT
>_Coding agent

Codex

Register SpendixAI once and use it from the Codex CLI, IDE extension or desktop app.

Set up Codex
Code editor

Cursor

Add SpendixAI to Cursor’s MCP settings and query your finances without leaving the editor.

Set up Cursor
Code editor

VS Code with GitHub Copilot

Add SpendixAI to VS Code and use it from Copilot chat in agent mode.

Set up VS Code with GitHub Copilot
Code editor

Windsurf (Devin Desktop)

Windsurf is now Devin Desktop. Add SpendixAI to its MCP config and sign in.

Set up Windsurf (Devin Desktop)
Coding agent

Antigravity (Gemini CLI)

Connect SpendixAI to Google’s Antigravity, or to Gemini CLI if you still use it.

Set up Antigravity (Gemini CLI)
Code editor

Zed

Add SpendixAI as a remote context server in Zed’s agent panel.

Set up Zed
Coding agent

Cline

Add SpendixAI as a remote server in Cline’s MCP panel, or through the Cline CLI.

Set up Cline
Coding agent

goose

Add SpendixAI to goose as a Streamable HTTP extension and sign in with OAuth.

Set up goose
Chat assistant

Perplexity

Add SpendixAI as a custom remote connector in Perplexity.

Set up Perplexity
Chat assistant

Mistral Vibe (Le Chat)

Add SpendixAI as a custom MCP connector in Mistral’s Vibe, formerly Le Chat.

Set up Mistral Vibe (Le Chat)
Coding agent

Warp

Add SpendixAI as an MCP server in Warp and ask the terminal agent about your spending.

Set up Warp
Coding agent

Junie

Connect SpendixAI to Junie CLI and ask JetBrains’ coding agent about your spending.

Set up Junie
Code editor

Kiro

Add SpendixAI to the Kiro IDE through its MCP config file and sign in from the browser.

Set up Kiro
Coding agent

GitHub Copilot CLI

One command adds SpendixAI to GitHub Copilot CLI so you can ask about your spending in the terminal.

Set up GitHub Copilot CLI
Coding agent

Kilo Code

Add SpendixAI to Kilo Code as a remote MCP server and sign in with OAuth from VS Code.

Set up Kilo Code
Coding agent

OpenCode

Add SpendixAI to OpenCode with one command and sign in with OAuth from the terminal.

Set up OpenCode
Coding agent

Amp

Store SpendixAI as a remote MCP definition in Amp, sign in once, and use it across Amp clients.

Set up Amp
Coding agent

Factory Droid

Add SpendixAI to Droid with one command and ask about your spending from the terminal.

Set up Factory Droid
Coding agent

Devin

Connect SpendixAI to Devin, in its web app or in Devin CLI, and query your records with OAuth sign-in.

Set up Devin
Coding agent

Replit

Add SpendixAI as a custom MCP server in Replit and let Replit Agent read your records after an OAuth sign-in.

Set up Replit
Coding agent

Lovable

Add SpendixAI to Lovable as a custom MCP connector and approve each change in a card.

Set up Lovable
Coding agent

Bolt

Add SpendixAI to Bolt as a custom MCP server using MCP OAuth, then check your spending from a chat.

Set up Bolt
Coding agent

v0

Try SpendixAI as a custom MCP server in v0; OAuth support is documented, but the form details are unconfirmed.

Set up v0
CrCoding agent

Crush

Add SpendixAI to Crush, Charm’s terminal coding agent, with a short config entry and OAuth turned on.

Set up Crush
Coding agent

Kimi Code CLI

Add SpendixAI to Kimi Code CLI with a config entry and /mcp-config login, and ask about spending in the terminal.

Set up Kimi Code CLI
Chat assistant

Microsoft Copilot Studio

Add SpendixAI as an MCP tool on a Copilot Studio agent so the agent can answer questions about your spending.

Set up Microsoft Copilot Studio
Chat assistant

Gemini app

Add SpendixAI to the Gemini app as a custom app and ask about your spending in a chat.

Set up Gemini app
Chat assistant

Raycast

Install SpendixAI as an MCP server in Raycast and ask about your spending from the launcher.

Set up Raycast
Chat assistant

Open WebUI

An admin adds SpendixAI to a self-hosted Open WebUI as an MCP tool server, and users sign in per chat.

Set up Open WebUI
Chat assistant

LibreChat

Add SpendixAI to a self-hosted LibreChat from MCP Settings or the librechat.yaml file.

Set up LibreChat
Chat assistant

LM Studio

Add SpendixAI to LM Studio's mcp.json and sign in through the browser, with a prompt before each tool runs.

Set up LM Studio
Chat assistant

Jan

Add SpendixAI to Jan through its /mcp console command and authenticate with a browser sign-in.

Set up Jan
TMChat assistant

TypingMind

Add SpendixAI as a custom connector in TypingMind and ask about your spending in a chat.

Set up TypingMind
Chat assistant

n8n

Use SpendixAI as an MCP tool inside n8n workflows and AI agents, with OAuth sign-in.

Set up n8n
Chat assistant

Zapier MCP Client

Connect SpendixAI to the Zapier MCP Client beta so your automations can use it as a tool.

Set up Zapier MCP Client
Chat assistant

Make

Add SpendixAI to a Make scenario through the MCP Client app and call its tools in an automation.

Set up Make
Chat assistant

Dify

Add SpendixAI as an MCP server in Dify and use it in agents and workflows.

Set up Dify
Chat assistant

Notion

Connect SpendixAI to a Notion Custom Agent as a custom MCP server on a Business or Enterprise plan.

Set up Notion
Chat assistant

Slack (Slackbot MCP client)

Let Slackbot use SpendixAI as an MCP server, with approval for every tool call.

Set up Slack (Slackbot MCP client)

Three steps, no keys to paste.

SpendixAI uses the standard sign-in flow, so you approve access on a consent screen instead of copying a secret into a config file.

Add the SpendixAI server

In your assistant, add the SpendixAI server address from Developer → MCP connections in your workspace.

Sign in and approve

Sign in to SpendixAI, choose the workspace and pick what the assistant may do.

Ask your first question

Start read only. Move to read and add, or full access, only when you need it.

You decide what an assistant can do.

The consent screen offers three presets. You can see connected assistants in your workspace and disconnect one whenever you like.

Read onlyRecommended

Ask about transactions, summaries and reports. Nothing can be added, changed or uploaded.

Read and add

Ask questions, add expenses and upload statements or receipts.

Full access

Everything above, plus categorisation rules and insights. Billing and member settings stay in the console.

Billing and member settings stay in the SpendixAI console and are not available to an assistant.

Any client that speaks MCP can connect.

SpendixAI works with clients that support MCP over Streamable HTTP with OAuth sign-in and dynamic client registration. If your assistant supports that, point it at the SpendixAI server address and follow the sign-in prompt.

Do I need to paste an API key?

No. The connection uses your SpendixAI sign-in and a consent screen, with dynamic client registration, so there is no secret to copy into a config file.

Can I keep the connection read only?

Yes. Read only is the recommended preset, and you can grant more access later or disconnect the assistant at any time from Developer → MCP connections.

Where do I find the server address?

In your SpendixAI workspace under Developer → MCP connections. Copy it from there, because that page always shows the address for the workspace you are signed in to.

Can an assistant change billing or members?

No. Those settings stay in the SpendixAI console.

Who sees my data?

The assistant’s provider processes the answers it gives, under its own terms. Connect only the workspaces you are comfortable discussing with it, and use the narrowest access that does the job.

Put your assistant to work on your own numbers.

Create your SpendixAI workspace, import a statement, and connect the AI you already use.