Global Logistics MCP
Remote MCP server that lets any AI agent buy shipping labels worldwide via AfterShip Shipping, using the merchant's own carrier accounts.
README
Global Logistics MCP
<!-- install-cta -->
Use it in 60 seconds
Paste this into your MCP client config (Claude Desktop, Cursor, Windsurf, or any MCP-capable agent):
{
"mcpServers": {
"global-logistics": {
"type": "http",
"url": "https://logi-global.wishpool.app/mcp"
}
}
}
Nothing to install. Credentials, when you need them, travel as HTTP headers on each request and are never stored — see the threat model.
Or run it yourself
Would you rather not send production credentials to a server you do not control? Deploy this identical code to your own account and point your agent at your own URL:
git clone https://github.com/junter1989k-ai/global-logistics-mcp && cd global-logistics-mcp && npx vercel --prod
MIT-licensed. Self-hosting removes us from the picture entirely, at no cost and with no loss of function.
Remote MCP server that lets any AI agent buy shipping labels worldwide via AfterShip Shipping (Postmen) — UPS, USPS, FedEx, DHL, Ninja Van, J&T and more — using the merchant's OWN bound carrier accounts. Stateless, bring-your-own API key, never stored.
- Endpoint:
https://logi-global.wishpool.app/mcp(MCP Streamable HTTP, stateless JSON) - Credential header:
x-aftership-key(your AfterShip key; forwarded upstream asas-api-key) - Environment header:
x-aftership-env=sandbox(default — real labels at zero cost) orproduction
The aggregation model (the honest core)
This server owns no carrier accounts. AfterShip is an aggregation layer: the merchant
binds their own carrier accounts (their contracts and rates) inside AfterShip
(admin.postmen.com) and passes that shipper_account_id on each call. This is exactly how an
agent reaches carriers that do not issue self-serve API keys — J&T, Ninja Van, Aramex —
the curve-around is the merchant's AfterShip-bound account. The sandbox can be tested with a fake
shipper account at zero cost.
Tools & capability matrix
| Tool | AfterShip endpoint | Supported | Notes |
|---|---|---|---|
quote_rates |
POST /postmen/v3/rates |
✅ yes | Compare live rates across the carriers on shipper_account_id. Read-only, buys nothing. Returns service_type + total_charge. |
create_label |
POST /postmen/v3/labels (pre-quoted via /rates) |
✅ yes | Buy a chosen service_type. Re-quotes the exact charge and applies owner policy guardrails before buying. Returns label_url + tracking_number. |
cancel_label |
POST /postmen/v3/cancel-labels |
✅ yes | Void an unused label by label_id. cancel_status: cancelled / pending / failed. Idempotent. |
query_tracking |
— | ❌ not supported | The Shipping (Postmen) API has no tracking endpoint. Tracking is a separate product: the AfterShip Tracking API. Returns a structured not_supported result (never errors). |
Addresses use ISO 3166-1 alpha-3 country codes (USA, GBR, JPN). Weight and dimension are
{value, unit} (unit kg|lb|oz|g and cm|in|m).
Owner policy guardrails
Set by the human owner in the MCP client config; the agent cannot relax them. The label charge is gated before any purchase.
x-agentpay-max-amount— hard cap; over it,POLICY_BLOCKED.x-agentpay-approval-above— over it, returns an unsigned draft (rate quoted, no label bought).x-agentpay-allowed-tools— tool allow-list.
Safety
Stateless translation layer: the AfterShip API key travels per-request in a header and is never stored; labels are generated and served by AfterShip/the carrier, never held here. Parcels flow sender ↔ carrier ↔ recipient directly.
Develop
npm run dev # local server on http://localhost:3242
npm test # e2e: protocol + validation + units + fake-key real 401 fingerprint
The e2e suite includes a live probe: a fake key really hits sandbox-api.aftership.com and
must receive the native 401 "The API key is invalid." fingerprint (proving the endpoint is alive).
Sister servers
US domestic labels via EasyPost — usa-logistics-mcp · Taiwan CVS pickup & home delivery — taiwan-logistics-mcp · local payments in 81 countries — mcp.wishpool.app · electronic-invoice servers across nine countries (Mexico CFDI, Brazil NF-e, Chile DTE, Peru CPE, India GST).
MIT.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.