MCP Server

Connect any MCP-compatible AI assistant to your lists

Overview

bundlist includes a built-in Model Context Protocol (MCP) server. MCP is an open standard that lets AI assistants read and write structured data on your behalf — without you copying and pasting anything between apps.

Once connected, your AI assistant can create lists, add and update items, retrieve your data by name or ID, and be notified when anything changes. Your lists become a shared, persistent workspace between you and your AI.

Connecting

The bundlist MCP endpoint is:

https://api.bundlist.app/mcp

It uses the streamable HTTP transport (HTTP + Server-Sent Events). Most MCP clients support this out of the box.

Step-by-step setup guides are available for:

• Claude.ai — connect via Customize → Connectors
• Claude Code — add via CLI or config file
• Gemini CLI — add via ~/.gemini/settings.json
• Cursor — add via Settings → Tools & MCP

Any client that implements the MCP 2025-03 specification should work. No API key or manual token setup is required — authentication is handled automatically via OAuth.

Authentication

bundlist uses standard OAuth 2.0 with PKCE. When you connect for the first time, your MCP client will open a browser window and prompt you to sign in with Google, Apple, or Facebook. After sign-in, your client stores the token and reconnects automatically.

You do not need to create an API key or configure anything manually. If your client supports Dynamic Client Registration (RFC 7591), the entire flow is automatic.

Available tools

The MCP server exposes 12 tools organized into three groups:

Read

• list_lists — return all your lists with field and item counts
• get_list — fetch a single list, optionally including all items and field values
• get_lists — fetch multiple lists by ID in a single call
• get_schema — return the field definitions for a list
• get_me — return your identity, tier, limits, and current usage
• export_data — download a full JSON snapshot of all your lists, fields, and items

Write

• create_list — create a new list, optionally with an initial set of fields
• rename_list — rename an existing list
• merge_list — create or update fields and items in a list in a single call; safe to call repeatedly (existing items are updated, not duplicated)

Delete

• delete_list — permanently delete a list and all its data
• delete_field — remove a field and its values from all items in a list
• delete_item — remove a single item from a list

What you can do

A few things to try once your AI assistant is connected:

Meal plan + grocery list in one go

Ask your assistant to plan a week of dinners, then turn the plan into a shopping list — organized by grocery section so you can move through the store efficiently.

Plan 5 weeknight dinners for a family of four, then create a grocery list in bundlist with the ingredients grouped by section (produce, dairy, meat, etc.)

Packing list for any trip

Describe your trip and let your assistant build a tailored packing list — categorized, realistic, and ready to check off in the app.

I'm taking two kids on a week-long beach trip in August. Build me a packing list in bundlist with categories for clothes, gear, toiletries, and entertainment

Gift ideas on demand

Describe someone and your budget — your assistant brainstorms ideas and adds them directly to your gift ideas list, ready to reference when you're shopping.

My daughter turns 10 next month and loves art and animals. Add 5 gift ideas under $40 to my gift ideas list in bundlist

Keep your watchlist current

Mark something as watched and get new recommendations added in the same conversation — your list stays up to date without you manually hunting for what to watch next.

I just finished Severance season 2. Mark it done in my watchlist and add 3 shows I'd probably enjoy based on what's already in the list

Build a workout schedule

Tell your assistant your goals, how many days a week you can train, and any equipment you have. It creates a structured plan as a list you can follow and update as you progress.

I want to build strength and have 4 days a week and a set of dumbbells at home. Create a workout schedule in bundlist with the day, muscle group, and exercises for each session

Meeting notes → action items

Paste your meeting notes and your assistant pulls out every action item — with owner and due date — and adds them to your list. No more manually re-reading notes to find what you committed to.

[paste meeting notes] Extract all the action items from these notes and add them to my action items list in bundlist with the owner and due date for each

Your AI preferences, everywhere

Keep a list of how you like your AI assistant to behave — tone, formatting rules, things to avoid. Load it at the start of any session and your preferences travel with you across every AI surface without copy-pasting.

Fetch my AI preferences list from bundlist and apply those rules to this session

Real-time notifications

The MCP server sends notifications/resources/updated events whenever your data changes. AI agents can subscribe to these to know when to re-fetch data without polling.

Troubleshooting

• Client can't connect — confirm your client supports streamable HTTP transport and the MCP 2025-03 spec. SSE-only or stdio-only clients are not supported.
• Authentication loop — try disconnecting and reconnecting the server in your client settings to restart the OAuth flow.
• Tool calls return 401 — your token may have expired. Reconnect the server to trigger a fresh sign-in.
• Items not appearing after create — check your account limits with get_me. Writes that exceed your tier limit are skipped and reported in the tool response.

Support

Having trouble connecting? Email support@bundlist.app and include your MCP client name and version.