# GET

**Kind:** Function

**Source:** [`atloria-monorepo/apps/web/src/app/p/[slugId]/pages/[slug]/route.ts`](https://github.com/sherkety/atloria/blob/main/atloria-monorepo/apps/web/src/app/p/[slugId]/pages/[slug]/route.ts#L17)

Same-origin markdown for one published page: /p/<slug>/pages/<page>.md

Every entry in llms.txt points at one of these, and until now every one of them 404'd into the
SPA shell — an agent following our own discovery index received 25 KB of HTML with a `200` and
a `text/html` content type, which is worse than a clean failure because nothing looks broken.
The API served the markdown correctly the whole time; nothing on the published origin routed
to it.

Proxied rather than redirected, for the same reason llms.txt is: an agent should never have to
follow a hop to a different origin, and the API attributes the read to the real crawler.

## Signature

```ts
async function GET(req: Request, ctx: { params: Promise<{ slugId: string; slug: string }> })
```

## Parameters

| Name | Type |
|---|---|
| `req` | `Request` |
| `ctx` | `{ params: Promise<{ slugId: string; slug: string }> }` |