# Angadi Labs

> Angadi is a Shopify app that lets fashion brands curate "Complete the Look"
> outfit recommendations. The brand decides what goes live: pairings are
> approved one by one by default, or, if the brand opts in, auto-published
> with one-tap removal. Angadi exposes a public, read-only feed so AI shopping
> agents can retrieve a brand's published outfit pairings for any product,
> with prices, availability, and canonical store links.

## Agent feed

- Endpoint: `GET https://angadi-complete-the-look.vercel.app/agent/v1/looks`
- Docs: [Agent feed documentation](https://angadilabs.com/docs/agent-feed)

### Request

Query parameters:

- `shop` (required): the store's myshopify domain, e.g. `example.myshopify.com`
- `handle` OR `product_id` (one required): the anchor product, as its Shopify
  product handle (e.g. `block-print-kurta`) or numeric Shopify product ID

### Response

JSON. `looks` is the list of brand-approved, currently-live outfits for the
anchor product. Each look has a stable `look_id`, a `type`
(`complete-the-look` = a full curated outfit, `style-it-with` = pairing
options for this product), and `pieces`. Each piece uses schema.org-compatible
fields: `name`, `url` (canonical storefront product URL), `image`, `price`,
`priceCurrency` (ISO-4217, omitted when unknown), and `availability`
(`https://schema.org/InStock` — out-of-stock pieces are omitted entirely,
matching the brand's storefront widget).

### Worked example

Request:

    GET https://angadi-complete-the-look.vercel.app/agent/v1/looks?shop=example.myshopify.com&handle=block-print-kurta

Response (abridged):

    {
      "version": "v1",
      "shop": "example.myshopify.com",
      "anchor": {
        "product_id": "gid://shopify/Product/1111",
        "handle": "block-print-kurta",
        "name": "Block Print Kurta",
        "url": "https://example.myshopify.com/products/block-print-kurta"
      },
      "look_count": 1,
      "looks": [
        {
          "look_id": "a1b2c3d4-...",
          "type": "complete-the-look",
          "pieces": [
            {
              "product_id": "gid://shopify/Product/2222",
              "name": "Palazzo Pants",
              "url": "https://example.myshopify.com/products/palazzo-pants?ref=angadi_ctlw&look_id=a1b2c3d4-...&source_product=gid%3A%2F%2Fshopify%2FProduct%2F1111&surface=agent_feed",
              "image": "https://cdn.shopify.com/.../palazzo.jpg",
              "price": "2099.00",
              "priceCurrency": "INR",
              "availability": "https://schema.org/InStock"
            }
          ]
        }
      ]
    }

### Notes

- Only brand-approved, live looks are returned. Nothing AI-generated is
  exposed before a human at the brand approves it.
- Responses are CDN-cached for 5-10 minutes. Requests are rate-limited per IP.
- The feed is off by default and enabled per brand: brands who want it
  switched on write to hello@angadilabs.com. Shops without the feed
  enabled (and unknown shops) return a minimal HTTP 404
  (`{"version":"v1","error":"not_found"}`). A
  handle or product ID that doesn't resolve returns 404
  `product_not_found`. A valid product with no approved pairings returns
  200 with an empty `looks` array.
- Product URLs carry the same attribution parameters the on-store widget uses;
  follow them as-is so the brand can see agent-driven engagement.

## More

- [Angadi Labs website](https://angadilabs.com)
- [Privacy policy](https://angadilabs.com/privacy)
- [Terms of service](https://angadilabs.com/terms)