Faslet MCP Server

The Faslet MCP server exposes Faslet sizing tools over MCP so AI assistants can resolve supported product pages and return size recommendations.

Endpoint

https://api.faslet.net/widget/mcp

The server uses MCP over HTTP with JSON-RPC 2.0 and currently reports:

• protocolVersion: 2024-11-05
• serverInfo.name: faslet-mcp

Example client configuration

For clients that support HTTP MCP servers, point them at the Faslet MCP endpoint directly:

{
  "servers": {
    "faslet/faslet": {
      "type": "http",
      "url": "https://api.faslet.net/widget/mcp"
    }
  }
}

Recommended tool flow

If you have a shopper-facing product page URL, start with get_size_from_product_url. That is the default tool and avoids asking the user for internal Faslet identifiers.

Only use get_size_from_product_info when shopId, brand, and productId are already known from trusted context or from get_product_info_from_product_url.

Tools

Tool	Use it for	Required arguments
get_size_from_product_url	Default sizing flow when you have a PDP URL. Resolves the product and then returns a recommendation when the store and product are supported.	productUrl
get_product_info_from_product_url	Resolve Faslet product info from a PDP URL before sizing.	productUrl
get_size_from_product_info	Get a size recommendation when Faslet product identifiers are already known.	shopId, brand, productId

Shared sizing inputs

Both sizing tools support these shopper inputs when available:

Field	Type	Notes
height	number	Height in centimeters
weight	number	Weight in kilograms
userId	string	Optional shopper identifier
userGender	string	male or female
age	number	Optional age
fit	string	tight, average, loose
waist	string	flat, average, round
hips	string	narrow, average, wide
chest	string	narrow, average, wide
footWidth	string	narrow, average, wide
braCupSize	string	Optional bra cup size
braBandSize	number	Optional bra band size
relativeSizingPreference	number	Optional sizing preference input
sizeLabels	string	Comma-separated size labels, for example S, M, L
experimentVariants	string	Optional experiment variant string

Response shape

Product discovery

get_product_info_from_product_url returns:

{
  "status": "resolved",
  "explanation": "Resolved product info from productUrl.",
  "productInfo": {
    "shopId": "faslet-demo",
    "brand": "Faslet Demo",
    "productId": "8390931513510",
    "productName": "Demo Product",
    "productUrl": "https://demo.faslet.app/products/demo-product"
  }
}

If the URL cannot be matched, the tool returns statuses such as invalid_input, store_not_supported, or product_not_supported.

Sizing tools

Sizing tools return content, structuredContent, and isError. The most useful payload is structuredContent:

{
  "status": "success",
  "explanation": "Recommended size: M",
  "missingInputs": [],
  "recommendation": {
    "label": "M"
  },
  "alternatives": [
    {
      "label": "M",
      "fit": {
        "waist": "perfect",
        "chest": "slightlyLoose"
      }
    }
  ],
  "warnings": [
    "userGender was inferred from product context."
  ],
  "metadata": {
    "shopId": "faslet-demo",
    "brand": "Faslet Demo",
    "productId": "8390931513510",
    "productGender": "male",
    "userGender": "male"
  }
}

Common sizing statuses:

Status	Meaning
success	Recommendation available
invalid_input	Missing or invalid sizing inputs. Check missingInputs.
no_recommendation	Product is supported, but no recommendation could be returned
backend_error	Unexpected internal failure
store_not_supported	The product URL belongs to a store that does not use Faslet sizing
product_not_supported	The store uses Faslet sizing, but that product is excluded

store_not_supported and product_not_supported are returned as normal tool results with recommendation: null, not as transport-level MCP errors.

Prompts

The server also exposes four prompts:

Prompt	Purpose
demo_basic_sizing	Example get_size_from_product_info call for one shopper profile
demo_preference_adjustments	Example sizing call using fit and body-shape preference enums
demo_url_discovery_sizing	Example flow that starts from a shopper-facing product URL
explain_measurement_inputs	Quick explanation of the enum-style sizing inputs

Example JSON-RPC calls

Initialize:

curl https://api.faslet.net/widget/mcp \
  -H 'content-type: application/json' \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "initialize"
  }'

Get a size recommendation from a product URL:

curl https://api.faslet.net/widget/mcp \
  -H 'content-type: application/json' \
  -d '{
    "jsonrpc": "2.0",
    "id": 2,
    "method": "tools/call",
    "params": {
      "name": "get_size_from_product_url",
      "arguments": {
        "productUrl": "https://demo.faslet.app/products/demo-product",
        "height": 176,
        "weight": 85
      }
    }
  }'

If the server needs more sizing information for a specific product type, it returns missingInputs so the client can ask a focused follow-up question.