Skip to main content
Glama
xentral

shiplabel-mcp

by xentral
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Remote

shiplabel-mcp

Carrier-agnostic shipping labels as a self-hostable MCP server. Build one shipment request, get a tracking number and a print-ready label back — the same way for DHL, DPD, UPS, FedEx, GLS, Sendcloud, Shipcloud and DHL Return. Run it yourself, connect it to Claude (or any MCP client), and create labels straight from a chat, a script, or your own agent.

Carrier-direct: no account with anyone is required to run this — you bring your own account with the carrier(s) you ship with.

Built and open-sourced by Xentral, the ERP for growing product businesses. This server is fully standalone and needs no Xentral account.

Don't want to self-host? The same engine is available ready-to-use, fully hosted, as the Carrier Kit in Xentral AgentOS — no server to run, no setup: agent.xentral.com/en/starter-kits.

Try it in 2 minutes

Looking around needs no carrier account — list_carriers and describe_carrier work out of the box:

pip install shiplabel-mcp        # or: uv pip install shiplabel-mcp
shiplabel carriers               # lists every carrier, no credentials needed

Your first real label — the fastest path is Sendcloud (self-serve API key, no per-carrier contract). Grab a public/secret key and a shipping-method id from the Sendcloud panel, then:

export SHIPLABEL_SENDCLOUD_PUBLIC_KEY="..."
export SHIPLABEL_SENDCLOUD_SECRET_KEY="..."
export SHIPLABEL_SENDCLOUD_METHOD_ID="8"      # a shipping method from your panel
shiplabel create --carrier sendcloud --from examples/sendcloud_request.json --out label.pdf

Prefer DHL? The DHL sandbox needs no production contract. Full setup for every carrier: per-carrier guides · configuration.

Related MCP server: royalmail-mcp

How it works

Carriers are data, not code. One generic engine executes a declarative JSON spec per carrier (endpoints, auth, a payload template, response paths). You build a single canonical shipment request (address + parcel + options); the engine maps it onto the carrier's API and normalizes the response to tracking number + base64 label. Adding or tweaking a carrier is a JSON file, not a code change.

Supported carriers

Carrier

Sandbox

What you need (your own account)

dhl — DHL Paket (DE)

✅

developer.dhl.com app + DHL business/GKP contract — see the note below

dhl_return — DHL Return (DE)

✅

DHL returns API key + receiver id

dpd

—

DPD business account (partner + cloud credentials)

ups

✅

UPS developer app + account number

fedex

✅

FedEx developer app + account number

gls

—

GLS business account

sendcloud

—

Sendcloud account (self-serve API key; aggregates PostNL, Swiss Post, Österr. Post, DPD, DHL…)

shipcloud

—

Shipcloud account (self-serve API key) — spec shipped, not yet exercised in tests

Bring your own carrier account. Every production carrier API requires a business/shipping account with that carrier. This project provides the integration; it does not include and cannot provide carrier credentials. The easiest self-serve entry points are the aggregators Sendcloud and Shipcloud.

Per-carrier setup & examples

Each guide has a concrete example: where to register, the exact env config, an example request, and the command to create a label.

  • DHL — includes a free sandbox quickstart

  • GLS

  • UPS — has a sandbox

  • Sendcloud — self-serve keys, easiest to start

  • DPD

  • FedEx, Shipcloud and DHL Return follow the same pattern — run describe_carrier <code> for their keys and see the configuration guide below.

→ Configuration guide — how credentials and options reach any carrier (env vars, TOML profiles, inline config, sandbox flags, adding your own carrier). Same mechanism for all of them.

⚠️ DHL needs your own credentials

This repo ships no DHL keys. To use DHL you need:

  1. your own app on developer.dhl.com (client id + secret) — free to register; sandbox works immediately;

  2. for production, additionally a DHL business-customer contract (Post & DHL Geschäftskundenportal, "GKP") with a customer/billing number. You don't get this "out of the box" — you register with DHL as a business customer.

To try it out, the sandbox is enough (public DHL test login, see docs/carriers/dhl.md). Without a GKP contract you cannot create real (production) labels — that's a DHL requirement, not a limit of this tool.

Quickstart

Option A — Docker

git clone https://github.com/xentral/shiplabel-mcp.git
cd shiplabel-mcp
cp .env.example .env        # fill in the carrier(s) you use
docker compose up           # HTTP MCP server on http://127.0.0.1:8000/mcp

Option B — local (Python 3.11+)

pip install shiplabel-mcp          # or: uv pip install shiplabel-mcp
cp .env.example .env               # and export/source it, or set env vars directly
shiplabel-mcp                      # stdio server (for Claude Desktop / Claude Code)
shiplabel-mcp --http               # or streamable HTTP on 127.0.0.1:8000

Connect it to an MCP client

The server exposes three tools: list_carriers, describe_carrier, create_label.

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "shiplabel": {
      "command": "shiplabel-mcp",
      "env": {
        "DHL_API_CLIENT_ID_SANDBOX": "your-dev-app-id",
        "DHL_API_CLIENT_SECRET_SANDBOX": "your-dev-app-secret",
        "SHIPLABEL_DHL_USERNAME": "your-gkp-user",
        "SHIPLABEL_DHL_PASSWORD": "your-gkp-password",
        "SHIPLABEL_DHL_ACCOUNTNUMBER": "your-billing-number",
        "SHIPLABEL_DHL_SANDBOX": "true"
      }
    }
  }
}

Claude Code

claude mcp add shiplabel \
  -e SHIPLABEL_DHL_SANDBOX=true \
  -e SHIPLABEL_DHL_USERNAME=... \
  -- shiplabel-mcp

HTTP mode

Start with shiplabel-mcp --http and point your client at http://127.0.0.1:8000/mcp (streamable HTTP transport).

Then just ask: "list the shipping carriers", "describe what dhl needs", "create a DHL label from Muster GmbH, Bonn to Erika Beispiel, Bonn, 1.5 kg."

Use it as a library or CLI

The MCP server is a thin wrapper over the shiplabel Python package, which you can also use directly:

from decimal import Decimal
from shiplabel import CanonicalShipmentRequest, CarrierSelection, Party, Parcel, create_label

req = CanonicalShipmentRequest(
    carrier=CarrierSelection(code="dhl", product="V01PAK"),
    sender=Party(name="Muster GmbH", street="Sträßchensweg", house_number="10",
                 postal_code="53113", city="Bonn", country="DE"),
    recipient=Party(name="Erika Beispiel", street="Kurt-Schumacher-Str.", house_number="20",
                    postal_code="53113", city="Bonn", country="DE"),
    parcels=[Parcel(id="p1", weight_kg=Decimal("1.5"))],
)
config = {"dhl_username": "...", "dhl_password": "...", "dhl_accountnumber": "...",
          "dhl_api_key": "...", "dhl_api_secret": "...", "dhl_sandbox": True}
result = create_label(config, req)
print(result.parcels[0].tracking_number)  # + result.parcels[0].label.data (base64 PDF)
shiplabel carriers                                  # list carriers
echo '{...}' | shiplabel create --carrier dhl --out label.pdf   # canonical request on stdin

See src/shiplabel/README.md for the full library / CLI reference and the canonical request shape.

Configuration

Copy .env.example and set only the carriers you use.

  • SHIPLABEL_<KEY> → the lowercase carrier config key <key> (e.g. SHIPLABEL_DHL_USERNAME → dhl_username).

  • DHL developer-app credentials are read from DHL_API_CLIENT_ID[_SANDBOX] / DHL_API_CLIENT_SECRET[_SANDBOX].

  • SHIPLABEL_CARRIERS_DIR — a directory of extra *.json specs to add or override carriers without forking.

Credentials can always also be passed inline per call (the MCP create_label config argument, or the library config dict) — inline wins over env.

See the configuration guide for TOML profiles, source precedence, sandbox flags, and the full canonical request shape.

Add a carrier

Drop a <code>.json spec into src/shiplabel/carriers/ (or a SHIPLABEL_CARRIERS_DIR). A spec has five parts — transport, auth, capabilities, request (a Jinja payload template), response. See src/shiplabel/README.md and dhl.json for a complete example. Modern REST/JSON carrier APIs fit the declarative model; carriers needing computed security (e.g. SOAP WSSE) are out of scope.

Security

  • Never commit credentials. .env, *.env (except .env.example) and carriers.toml are git-ignored.

  • Labels are returned as base64 blobs; the CLI writes them to disk only where you ask. Generated *.pdf/*.zpl/*.png are git-ignored.

  • Report vulnerabilities per SECURITY.md.

Development

uv venv && source .venv/bin/activate
uv pip install -e ".[dev]"
python -m pytest        # transport is mocked — no live carrier calls
ruff check .

License

MIT © Xentral ERP Software GmbH.

Available Tools

3 tools
create_labelA

Create a shipping label from a canonical shipment request.

request is a canonical shipment request: {carrier: {code, product}, sender, recipient, parcels: [{id, weight_kg, dimensions_cm}], references, label: {format}}. Addresses need name/street/postal_code/city/country.

carrier overrides request.carrier.code when set. config supplies carrier credential keys (e.g. dhl_username) merged over the environment; use it for ad-hoc testing without setting SHIPLABEL_* env vars. include_label=false returns a compact tracking-only reply without the base64 label.

Returns the shipment number and, per parcel, the tracking number/URL, label format and (unless disabled) the base64-encoded label.

ParametersJSON Schema
NameRequiredDescriptionDefault
configNo
carrierNo
requestYes
include_labelNo

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4.4/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description discloses behavioral traits: input structure, override logic, optional output suppression, and return format. It mentions that config is for testing, implying normal credential loading from environment. It does not mention costs or destructive nature, but the behavior is well-explained.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured: purpose first, then request details, then optional parameters and return value. It is somewhat verbose but every sentence adds value. Front-loading works well.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the complexity (4 params, nested objects, output schema exists), the description covers input structure, overrides, credential handling, and return format. It does not detail all sub-fields but provides sufficient context for an agent to use the tool correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, but the description fully compensates by detailing the nested request structure, address fields, carrier/product, parcels, references, label format, and the purpose of carrier, config, and include_label. This adds significant meaning beyond the bare schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states 'Create a shipping label from a canonical shipment request,' specifying the verb, resource, and primary action. It distinguishes itself from siblings (describe_carrier, list_carriers) which are about carrier info, not label creation.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description explains when to use optional parameters like 'carrier' (overrides request value) and 'config' (for ad-hoc testing without env vars). It does not explicitly state when not to use the tool or compare to alternatives, but the sibling tools are clearly different in purpose.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

describe_carrierA

Describe the canonical shipment request and, if carrier is given, that carrier's required config keys, supported services and label formats.

Call this before create_label to learn exactly what to pass.

ParametersJSON Schema
NameRequiredDescriptionDefault
carrierNo

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4.4/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations provided, so description carries full burden. It discloses what is returned (required config keys, supported services, label formats) and implies read-only behavior. Does not mention side effects, auth needs, or errors, but is sufficient for a describe tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two sentences, no wasted words. Purpose and usage are front-loaded. Highly concise.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Output schema exists, so description need not detail return format. It covers what info is returned and usage hint. Lacks mention of alternative sibling 'list_carriers', but overall complete for a simple describe tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema has 0% description coverage for parameter, but description explains that if 'carrier' is given, carrier-specific details are returned; otherwise, canonical request. Adds meaning beyond bare schema, though could clarify null behavior more explicitly.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Description clearly states it describes the canonical shipment request and carrier-specific details if carrier is given. Verb 'describe' and resource 'shipment request and carrier config' are specific. Distinguishes from sibling 'create_label' by advising to call before it.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly says 'Call this before `create_label` to learn exactly what to pass,' providing clear when to use. Does not mention when not to use or alternatives, but context with 'list_carriers' suggests this is for detailed info.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

list_carriersA

List the shipping carriers this server can create labels for.

ParametersJSON Schema
NameRequiredDescriptionDefault

No parameters

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description must disclose behavioral traits. It states the tool lists carriers, but does not mention ordering, filtering, side effects, or scope. Since the tool has no parameters and a likely output schema, the transparency is adequate but minimal.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single clear sentence with no unnecessary words. Every part contributes meaning, making it highly concise and well-structured.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool has no parameters and an output schema is provided, the description sufficiently explains the tool's purpose. It could mention that it returns all carriers without filtering, but the context is already complete for a simple listing tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has zero parameters with 100% coverage. The description adds no parameter info because none is needed; it implicitly indicates no parameters are required. This is a baseline 4 for no-parameter tools.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's function: to list shipping carriers that the server can create labels for. It uses a specific verb ('List') and resource ('carriers'), and distinguishes itself from sibling tools (create_label, describe_carrier) by its focus on enumeration rather than creation or details.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description does not explicitly state when to use this tool versus alternatives like describe_carrier or create_label. Usage is implied by the tool's name and purpose, but no direct guidance is given, which leaves the agent to infer from context.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

  1. 3 tool updatesv0.1.0
    • First observedcreate_label
    • First observeddescribe_carrier
    • First observedlist_carriers
Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A4.3/5.0

Scored across 3 tools

Disambiguation5/5

Each tool has a distinct purpose: list_carriers discovers carriers, describe_carrier explains requirements, create_label performs the main action. No functional overlap.

Naming Consistency5/5

All tools follow a consistent verb_noun pattern with lowercase and underscores (list_carriers, describe_carrier, create_label), making the API predictable.

Tool Count4/5

Three tools is slightly minimal but appropriate for a focused shipping label MCP. It covers discovery, guidance, and creation without extra fluff.

Completeness4/5

The core workflow (list carriers, describe requirements, create label) is fully covered. Missing operations like label retrieval or cancellation are minor gaps given the server's purpose.

Maintenance

ActivityStale
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers