Freight Visibility Inside Your AI Tool

What is MCP and why it matters for freight

Model Context Protocol is an open standard that lets AI tools call external services natively. Instead of your AI tool writing curl commands or HTTP requests, it calls Warp tools directly. The AI sees what tools are available, understands their parameters, and executes freight operations as naturally as it reads files or writes code.

For freight, this means your AI coding assistant becomes a freight visibility tool. It doesn't need to parse API documentation or construct JSON payloads. The MCP server exposes tracking, shipment search, document retrieval, event history, and invoice lookup as typed tool calls with structured inputs and outputs. The AI fills in the parameters from your natural language description and executes.

This is a step change from API integration. An API requires your AI to read docs, write code, handle errors, and parse responses. An MCP server hands the AI a menu of freight operations with typed schemas. The AI calls them like any other tool.

Two ways to connect

{
  "mcpServers": {
    "warp": {
      "disabled": false,
      "headers": {
        "x-api-key": "YOUR_WARP_API_KEY"
      },
      "serverUrl": "https://server-mcp.wearewarp.com/sse"
    }
  }
}

{
  "mcpServers": {
    "warp-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "wearewarp-nodejs-mcp"
      ],
      "env": {
        "WARP_API_KEY": "YOUR_WARP_API_KEY"
      }
    }
  }
}

Both methods require a Warp API key. Get one from customer.wearewarp.comunder Settings > API Keys.

What you can do through the MCP server

• -Track shipments by tracking number with real time status, location, and ETA.
• -List and filter all your shipments by order number, date range, or reference number.
• -Pull BOLs, PODs, delivery photos, and signatures for any order.
• -View the full event timeline for any shipment: pickup, transit, cross dock, delivery.
• -Audit invoices with full cost breakdowns: transit, fuel, service options, grand total.
• -Chain tools together: look up an order by name, then pull its documents in one conversation.
• -Quoting and booking tools are on the roadmap. Use the REST API or CLI for those today.

What it looks like

This is a real conversation with Claude Code connected to the Warp MCP server. You ask about your freight. The AI calls Warp tools and gets you answers.

You: "Where is shipment WARP-4829173?"

Claude Code calls warp.tracking...

Status: IN TRANSIT
  Current Location: Nashville, TN
  Last Scan: April 7, 6:42 PM
  ETA: April 9, 2:00 PM
  Temperature: 37F (within range)

You: "Show me the full event history."

Claude Code calls warp.get-shipment-events...

  April 7, 10:15 AM  Picked up in Atlanta, GA
  April 7, 2:30 PM   Departed ATL cross-dock
  April 7, 6:42 PM   In transit, Nashville, TN

You: "Pull the BOL and POD for that order."

Claude Code calls warp.get-documents...

  BOL: WARP-BOL-4829173.pdf (downloaded)
  POD: Signature + 2 delivery photos attached

You: "What was the final invoice?"

Claude Code calls warp.get-invoice...

  Transit: $1,247.00
  Fuel: $0.00 (all inclusive)
  Grand Total: $1,247.00
  Status: PAID

MCP server vs raw API

MCP use cases

Operations teams: freight visibility without portals

Your operations team checks shipment status 50 times a day. Each check means logging into a portal, searching by tracking number, clicking through tabs. With the MCP server connected to Claude Code or Claude Desktop, they ask in natural language: "Where is the Dallas shipment?" and get tracking, ETA, and last scan event instantly. No portal. No tab switching. No remembering which system has the data.

Customer service: answer freight questions in seconds

A customer calls asking about their delivery. Instead of putting them on hold while you search three systems, you ask your AI tool: "Pull tracking and the POD for order 78234." The MCP server calls get-shipments to find the order, then tracking for status, then get-documents for the proof of delivery photo. You have the answer before the customer finishes explaining the problem.

Invoice audit: catch discrepancies automatically

Tell your AI tool: "Pull invoices for all shipments from last month and flag anything over $2,000." The MCP server calls get-shipments filtered by date, then get-invoice for each order. The AI aggregates the data, flags outliers, and gives you a summary. No spreadsheet. No manual reconciliation. Every invoice checked in one conversation.

Exception detection: find problems before your team does

Ask: "Check all in-transit shipments and flag any that look late." The MCP server pulls your active shipments via get-shipments, then calls get-shipment-events for each one. The AI compares event timestamps against expected transit times and flags shipments where pickup was late, a scan is missing, or dwell time at a cross-dock is abnormal. You catch exceptions at 8am instead of when the customer calls at 4pm.

Document retrieval: BOLs and PODs on demand

"Pull the BOL for shipment WARP-4829173." One sentence. The MCP server calls get-documents and returns the PDF. No portal login. No digging through email. Works for BOLs, PODs, delivery photos, and e-signatures. Your accounting team can pull every document they need for month-end close through natural language.

Developers: build freight features with real data

Building a shipment tracking dashboard? Connect the MCP server in Cursor. Ask it to pull your real shipments. See the actual data shapes. Write the component that renders them. Test against live data without a separate API playground. The MCP server gives your AI code editor the same freight data your production code will consume.

Reporting: aggregate freight data through conversation

"How many shipments did we complete last week? Total invoiced amount? Average transit time?" The MCP server pulls the data. The AI does the math. You get a report in 30 seconds that would take an analyst 2 hours to build from portal exports. Chain it with follow-ups: "Which lanes had the highest cost per pallet?" or "Show me all shipments that missed their delivery window."

Compatible AI tools

• -Claude Code: Anthropic terminal AI. Full MCP support via SSE and stdio.
• -Claude Desktop: Anthropic desktop app. SSE and stdio support.
• -Cursor: AI code editor. MCP support in settings.
• -Windsurf: AI code editor with MCP integration.
• -Cline: VS Code extension with MCP support.
• -Any MCP compatible client: The protocol is open. If your tool supports MCP, it works with Warp.

The Warp network behind the MCP server

• -20,000+ local 3rd-party carriers dispatched through the Warp driver app.
• -9,000+ box trucks and cargo vans in network.
• -50+ Warp operated cross-dock facilities.
• -1,400+ active LTL lanes with all inclusive per pallet pricing.
• -Live GPS tracking, scan events, proof of delivery photos, e-signatures on every shipment.
• -Our AI backbone, Orbit, monitors every load for exceptions before your team has to chase them.
• -All inclusive pricing with no fuel surcharges or hidden fees.

Frequently asked questions

Related

• -Connect Warp MCP to Claude Code
• -Connect Warp MCP to Cursor
• -Ship Freight with Claude Code
• -AI Freight Agent
• -Programmable Freight Infrastructure
• -Vibe Coding Freight