Skip to content

Stacksync × Supabase Miami: AI-Native OperationsOct 13Reserve your spot

New

Stacksync MCP is here. Give Claude, Cursor, and any MCP client governed access to your syncs, workflows, and connections.

Set up MCP
Data warehouse · Two-way sync platform

MotherDuck connector for two-way sync.

Keep MotherDuck and your business systems aligned with Stacksync. Sync Tables (confirm coverage) within the connector’s object and field requirements, with two-way sync for eligible records.

Set this up from Claude or ChatGPT
  • Configure field mappings and sync direction
  • Inspect and resolve record errors in one dashboard
Setup guide: how Stacksync connects to MotherDuck

Adopted by fast-scaling companies moving mission-critical data in real time

Case study
Migrated from MuleSoft
Case study
Migrated from Celigo
Migrated from Heroku Connect
Migrated from Matillion
Case study
Migrated from Fivetran
Case study
Migrated from Celigo
Record types

MotherDuck objects and coverage.

Start with the records your workflow needs. Object-specific permissions and operation limits are shown alongside the available coverage.

Tables (confirm coverage)
Check object and field eligibility in the setup guide.

Read the MotherDuck object and setup guide

Connection setup

MotherDuck connection requirements.

Check the account access and change-detection requirements for your environment. Your system administrator can help confirm permissions before the first sync.

Connection method
SQL through DuckDB clients and drivers using a MotherDuck (md:) connection
Account access
Access token created in MotherDuck (Settings > General > Create Token), pasted into Stacksync; database name and schema configurable if not using defaults
How changes are detected
The Stacksync guide does not specify the change-detection mechanism. Confirm it for the selected objects.

Sync directions

  • Read data Documented
  • Write data Documented
MotherDuckStacksync setup guide
Setup prerequisites for your administrator
  • Generate a MotherDuck access token in Settings > General and enter it in the Stacksync connection form.
  • Specify the intended database and schema and confirm read/write coverage for the selected tables.

Documentation checked September 15, 2026.

Setup guides and technical references

Integration design

Match records and fields for your MotherDuck integration.

Choose the dataset grain, stable business key, and model refresh schedule. If results update an application, restrict writes to fields owned by the analytical process.

Use your account’s actual field names when applying these matching and dependency checks.

Explore record matching and field-mapping checks
MotherDuckSource record IDKeep the account and record type
Record matchingSource ID ↔ destination IDLink the same record across systems
Selected destinationDestination record IDPreserve its local key and rules
Conceptual identity model. Keep source and destination IDs linked so an update reaches the right record. Confirm the supported sync direction separately.
Match the same record in both systems
Choose a stable logical row key and state the table’s business grain. A changing analytical result should update the intended destination record instead of creating a new one on every refresh.
Load related records in the right order
Document the query or upstream process that populates the selected table. Measure its refresh age separately from the transfer into another system.

Mapping checks by record type

Tables (confirm coverage)

Check the object requirements below

Identity
Define a stable key for each table row or document and retain the external application key separately. Confirm the selected connector’s exact key requirements before altering a schema.
Fields to consider
Local record key · External application ID · Required business fields · Parent record references
Related records to resolve first
Document the query or upstream process that populates the selected table. Measure its refresh age separately from the transfer into another system.
Test before going live
Test repeated delivery, a changed external value, a missing parent, an incompatible type, and a schema change using a representative record.

Download the field-mapping workbook (CSV) to capture the actual API fields, matching keys, owners, and test results. Use it as you build with your team or a Stacksync engineer.

Setup and validation

Validate the MotherDuck integration in four steps.

The destination can trace a value to its source dataset and computation time, and a model refresh does not create duplicate operational records.

View the four setup tests and expected results

  1. Verify MotherDuck connectivity and database privileges

    Use the documented setup from the integration network to the intended host, database, and schema. Have the administrator demonstrate the permissions required for both the first read and later change capture.

    Expected result: The approved role can access the selected data and capture later changes without depending on an administrator’s personal session.

  2. Validate a table update and its source freshness

    Use the authorized MotherDuck token with the intended database and schema. Change a test value and observe its permitted destination.

    Expected result: The account can access the chosen table, and the measured delay includes both dataset refresh and synchronization.

  3. Test writes in the directions you need

    Select one eligible MotherDuck object and writable field. Test the required create/update direction, an invalid value, and a repeated delivery. Check object-specific restrictions before enabling the return path.

    Expected result: The selected operation is confirmed, addresses the intended record, and exposes rejected values for repair.

  4. Test recovery and name the support owner

    Interrupt a non-production transfer, restore the connection, and reconcile the selected MotherDuck records. When using Stacksync’s Retry to sync action, check the latest source value: the retry re-reads current source state rather than replaying an immutable historical event.

    Expected result: The records have the expected current values, repeated delivery creates no duplicate business record, and your team knows who handles unresolved errors.

Use the production-readiness checklist to record test results, assign support owners, and agree on when to go live.

Troubleshooting

Troubleshoot your MotherDuck integration.

Diagnose record matching, delayed changes, and rejected operations
The connection works but the expected data is absent
Verify the database/schema selection and token access. Confirm table eligibility, writable columns, and detection interval for the selected configuration.
The first load looks correct but later results differ
The Stacksync guide does not specify the change-detection mechanism. Confirm it for the selected objects. Measure the original source change, its capture, and the destination observation separately. Confirm whether the affected object is incremental, historical-only, or subject to a special refresh cycle.
One operation succeeds while another is rejected
Check the specific MotherDuck object, field permissions, required references, and allowed operation. A successful read does not establish that a create, update, deletion, or business action is available.

Review retry, replay, and recovery behavior before repeating an operation that may already have succeeded.

Review your MotherDuck workflow with an engineer

MotherDuck integration guides

FAQ

MotherDuck connector FAQ

How should I plan an integration with MotherDuck?

Specify the intended database, schema, table keys, and measured refresh behavior. Stacksync supports two-way sync for eligible objects. Start with one workflow, agree on which fields each system owns, then test initial data, later changes, and recovery before expanding.

Can the MotherDuck connector write changes back?

Stacksync supports reads and writes for eligible MotherDuck objects. Check the selected fields, account permissions, and the other system’s connector. Use the authorized MotherDuck token with the intended database and schema. Change a test value and observe its permitted destination.

Which identifiers should I preserve for MotherDuck data?

Choose a stable logical row key and state the table’s business grain. A changing analytical result should update the intended destination record instead of creating a new one on every refresh.

Which related records should I load first for MotherDuck?

Document the query or upstream process that populates the selected table. Measure its refresh age separately from the transfer into another system.

What should I check if MotherDuck data is missing or delayed?

Verify the database/schema selection and token access. Confirm table eligibility, writable columns, and detection interval for the selected configuration.

What will we cover in a MotherDuck integration demo?

Start with one workflow and a sample record from Tables (confirm coverage). We can review how to identify the record, connect related data, set the update direction, and test recovery. Include the person who owns the source system and anyone who will maintain the integration; you do not need a finished requirements document.

AI assistant

Set up MotherDuck from Claude or ChatGPT

Connect the Stacksync MCP server to your AI assistant, then ask for the sync in plain language. You only sign in; there is no code to write and no API key to paste.

1. Add the Stacksync MCP server

Server URL

https://mcp.stacksync.com/mcp

In Claude

  1. Open Customize → Connectors, select + Add, then Add custom connector.
  2. Name it Stacksync and enter https://mcp.stacksync.com/mcp.
  3. Select Connect, sign in to Stacksync and choose your workspace.

In ChatGPT

  1. Open Settings → Apps → Advanced settings and turn on Developer mode.
  2. Select Create app, name it Stacksync, enter https://mcp.stacksync.com/mcp and choose OAuth.
  3. Sign in to Stacksync, then enable Stacksync from the tools menu in a chat.
Set up Codex or Claude Code

Codex

codex mcp add stacksync --url https://mcp.stacksync.com/mcp
codex mcp login stacksync

Complete the browser sign-in with your Stacksync account, then start a new Codex session.

Claude Code

claude mcp add --transport http stacksync https://mcp.stacksync.com/mcp

Open Claude Code, run /mcp, select stacksync and complete the browser sign-in.

Then ask: “List my Stacksync workspaces.” Sign-in connects your assistant; it does not create a sync yet.

Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI and other assistants: setup steps in the docs.

2. Ask your assistant

Create a sync between my MotherDuck and another app or database.

Also from the same chat

You can also ask for workflows and AI agents (Genies) that use MotherDuck. The assistant checks what each app supports, for example:

  • “Build a workflow that messages our team in Slack when a MotherDuck record changes.”
  • “Create a Genie that answers questions about our MotherDuck data in Slack.”

What happens next

  • Signs you in with your Stacksync login (OAuth). No API key is pasted into the chat.
  • Gives you a secure sign-in link for each app it connects; you open it and sign in to that app.
  • Builds the sync with your permissions in Stacksync, nothing more. Turn on Ask before every change in the MCP settings to approve each change first.

New to Stacksync? Create an account first, then sign in from your assistant.

Explore MotherDuck connections

Choose the other system in your workflow. Each guide explains the connection options, available operations, field mapping, and setup checks for that pair.

197 integration guides

CRM · 30

ERP · 27

Database · 24

Data warehouse · 8

Business productivity · 34

Accounting and finance · 14

AI · 4

Analytics · 6

Communications · 5

Developer tools · 9

E-commerce · 6

Human resources · 12

Marketing · 8

Security and identity · 6

Storage · 4

SECURITY

Security teams trust Stacksync

As a data company, we understand the importance of keeping your data secure. Stacksync is built with security best practices to keep your data safe at every layer, and is DPF-certified for US, EU, UK and CH data transfers.

Learn more about security
SOC 2 Type II
ISO 27001
HIPAA BAA
GDPR
CCPA
DPF US-EU-UK-CH
→ SECURITY WITH BENEFITS

SSO & SCIM

Let your users access Stacksync from your centralized user management systems. Works with Okta, Azure, Google SSO and more.

Alerts

Immediately get alerted about record syncing issues over email, Slack, PagerDuty and WhatsApp. Resolve issues from a centralized dashboard with retry and revert options.

Secure connection options

Securely connects to your systems with:

OAuth 2SSH TunnellingSSL certificatesIP WhitelistingVPN gatewayVPC peeringand more
Coworkers laughing in front of a laptop in a casual office setting

Connect the systems your business runs on.
Build your next workflow with Stacksync.

Book a demo Get started