About MCP
The Model Context Protocol (MCP) is an open standard for connecting AI agents to the tools and data they work with. A service runs an MCP server that describes what it can do as a set of tools, and any MCP client, such as Claude, ChatGPT, Cursor or VS Code with GitHub Copilot, can call those tools on your behalf. You ask in plain language, and the agent works out which tools to call.
Why connect an AI agent to Zippendo?
Zippendo is a shipping platform for webshops: orders come in from your shop, shipping rules book them with your own carrier agreements, and labels print in the warehouse. Zippendo runs its own MCP server at `https://api.zippendo.com/mcp`, so the agent you already use can work with the same data you see in the app.
Ask questions about your shipping
"How many shipments are still pending, and which carriers are they on?" or "Which orders from the last 24 hours don't have a shipment yet?" The agent reads your orders and shipments and answers without you building a report or exporting a spreadsheet.
Let the agent do the routine work
With write access, the agent can create a shipment for an order, split it into two parcels and send both, or add a shipping rule that routes heavy orders to a particular carrier. Every call goes through the real Zippendo API, with the same validation, plan limits and checks as when you click in the app.
Build your own integration faster
The server also lets the agent read Zippendo's documentation, the full API reference and the SDK reference for TypeScript, Python, PHP, Go and Ruby. Ask for a script that creates a shipment in your language, and the agent writes it against the official SDK instead of guessing.
Safe by design
The agent acts as you, inside one organisation, and only with the permissions you approved on the consent screen. It can never perform Zippendo admin actions, and some sensitive operations are left out of the tool catalogue entirely, whatever permissions you grant.
What you can do with MCP and Zippendo
- Shipments: list, get and track shipments, fetch labels and documents, and create, split, send and return
- Orders: read synced orders, update them and create new ones
- Carriers: list connected carriers, their products and drop-off points, and connect and configure carriers
- Addresses and shipping rules: read and manage sender and return addresses and your shipping rules
- Automations, printers and webhooks: create, edit and run automations, register and approve printers, and manage webhooks
- Order channels, integrations and return portals: read and manage connected channels, integrations and return portals
- Documentation tools: `search_documentation`, `fetch_documentation` and `get_sdk_reference`, available without any scope
- Scopes you choose: `read:` and `write:` per resource, or `read:all` and `write:all`
- OAuth 2.1 login: no token to copy, and access refreshes on its own
- Same tools as Zippy: the MCP server shares its tool catalogue with Zippy, the assistant inside Zippendo
How do I connect an AI agent to Zippendo?
Your agent works with the data already in Zippendo, so connect your shop first. There are ready-made integrations for Shopify and WooCommerce, and any other system can send orders through the API. Add the carriers you ship with, for example GLS, PostNord or DAO. If you want events pushed to other tools rather than an agent asking for them, look at Slack, Zapier or Make. See all integrations.
Add `https://api.zippendo.com/mcp` to your client as a custom MCP connector. The client opens the Zippendo login in your browser, where you pick the organisation and approve the permissions. Connect your agent has step-by-step setup for Claude Code, Claude Desktop, Claude on the web, Cursor, VS Code, Windsurf, ChatGPT, Codex CLI and Gemini CLI, and Tools & scopes lists what each permission allows. Connected agents show up under Apps & integrations, where you can disconnect them.
MCP integration FAQ
Start automating with MCP
Add the connector URL to your agent, approve the permissions and ask your first question about your shipping.
