Incremental and resumable captures

A tori archive is meant to be kept, not captured once and forgotten. Re-running is cheap because tori knows what it already holds.

Fetching only what is new

tori add (alias tori update) re-captures an existing target but fetches only what is newer than the newest tweet already on disk, then re-renders the views:

tori add karpathy

Under the hood, an add reads the newest tweet id in the repository and uses it as the floor for the new fetch, so it pulls the gap since your last capture and nothing more. The capture summary shows the delta as (+N new). tori archive on an existing repo does the same incremental fetch; add is just the clearer name for a re-run.

You can set the floor by hand with --since-id, or cap the top with --until-id, when you want a specific id range.

Resuming an interrupted run

Every record is written to disk the instant it arrives, not buffered for the end. That means an interrupted run is never wasted:

• Press Ctrl-C and tori stops, keeping every record it already wrote. It exits with code 130.
• Hit the search rate limit on a long --by-month walk and tori exits 5 with the partial archive intact.

Either way, run the same command again (or tori add) and it continues from what is on disk, fetching only the rest.

Forcing a clean recapture

To ignore the held state and recapture from scratch (for example, to pull in edits to tweets you already have), use --force:

tori add karpathy --force

This refetches the full window rather than just the new tail, overwriting the records it finds again.

Reproducible output

tori's output is deterministic by design: record paths and media filenames are pure functions of their content, and the manifest's record-bearing fields are sorted. The one wall-clock value is the capture stamp. Pin it with --date to make a run byte-for-byte reproducible:

tori archive karpathy --guest --date 2025-06-17T00:00:00Z

The same --date is available on tori render to fix the footer stamp when you rebuild the views.

Previewing without fetching

--dry-run prints what a capture would target without touching the network:

tori archive karpathy --by-month --dry-run

A self-contained, movable archive

The repository is fully self-contained. Records, media, views, CSS, and the manifest all live under the one root directory, with every internal reference written as a relative path. Move the folder to another disk or machine, open index.html, and it still works with the network unplugged. To browse it over HTTP, point tori serve at it:

tori serve $HOME/data/tori/x/karpathy

Next

• The CLI reference lists every flag.
• Repository layout maps what lands on disk.