The Wayback Machine

The wayback group (alias wb) works with the Wayback Machine: the web-capture side of the Internet Archive, addressed by URL and timestamp rather than by item identifier.

The closest snapshot

archive wayback available example.com
archive wayback available example.com -t 2010

available asks the Availability API for the capture nearest a timestamp (-t, a full or partial YYYYMMDDhhmmss). It returns the snapshot's timestamp, HTTP status, and replay URL.

The full capture history

archive wayback list example.com -n 10
archive wayback list example.com --from 2010 --to 2012 --status 200
archive wayback cdx 'example.com/*' --match-type prefix --collapse digest

list (alias cdx) reads the CDX server, which returns one row per capture: timestamp, original URL, MIME type, status, digest, and length. Narrow it with --from/--to, --status, --mime, a raw --filter, or --collapse to fold adjacent duplicate rows on a field.

The CDX server is aggressively rate-limited by the Archive. archive throttles and retries with backoff automatically, but a busy moment can still return 429s; raise --rate and --retries and try again.

Fetching a snapshot

archive wayback get example.com -t 2010 --text     # readable text
archive wayback get example.com -t 2010 --links     # the page's hyperlinks
archive wayback get example.com -t 2010 --raw > page.html  # original bytes
archive wayback get example.com -t 2010 -o page.html        # write to a file

get resolves the closest snapshot (or the one at -t), then fetches the original archived bytes. --text extracts readable text, --links lists the hyperlinks (great with -o url), and the default is the raw archived HTML.

Saving a fresh capture

archive wayback save https://example.com/

Anonymously this is a fire-and-forget request to Save Page Now. With --outlinks or --screenshot it uses the authenticated SPN2 API (which needs credentials) and, with --wait, polls the capture job to completion:

archive wayback save https://example.com/ --outlinks --wait