Arquivo.pt

factory

providers.arquivo()

index

CDX

reads bodies

yes

provider=all

included

Load it

import { createArchive, providers } from "@agntn/archives";

const archive = createArchive(providers.arquivo());

No credentials, no extra options beyond the shared ones.

What it lists

snapshots queries arquivo.pt/wayback/cdx. A domain is treated as a prefix; a URL is looked up as given. Every row carries status, mime, digest and length in _meta, so two captures with the same digest are the same body even before you read them. The from and to window is applied in the query.

Arquivo.pt has archived far more than Portuguese sites: it holds a broad crawl of the early web and answers quickly, which is why it often contributes most of a merged listing.

Reading bodies

Bodies come from noFrame/replay, the raw playback path that serves the archived response without the archive's frame. Timestamp resolution is the shared rule: newest at or before, else closest after.

Gotchas

• Listings include captures with user:pass@ URLs and www. variants as separate originals. Match on digest when you want distinct bodies.
• The length field is a string, as the CDX API sends it.

Where it lives

src/providers/arquivo.ts.