Delivering & the ArchivesSpace Integration

Narrative drafts

The tab opens with drafts. A collection-level scope-and-content note, and a biographical note when the batch has a dominant creator. Drafts are grounded strictly in your documents and saved curation, no outside facts, and are advisory. Edit them, copy them into your finding aid, or ignore them. Nothing is stored and nothing sends automatically.

Readiness checks

Before anything delivers, readiness findings tell you what would go wrong.

• Undated documents. ArchivesSpace rejects EAD imports without a date, so these are called out with the affected documents named
• Untitled documents
• Unresolved sensitivity flags. Hold flagged records back from publication until cleared. Direct sends deliver them unpublished with a drafted restriction
• No access points selected, when a repository connection exists. Selection is opt-in, so an empty selection sends nothing
• Unresolved reconciliation conflicts
• No language of description on the institution profile. A note, not a warning. Exports and sends record English until you set it under Settings.

Findings are advisory. They tell you what to fix and where, and they never block you silently.

Sending to ArchivesSpace

Acervo drives ArchivesSpace's standard API. Nothing to install on your instance, no plugin. Connect once under Settings, then Integrations, then ArchivesSpace. You provide the backend API URL, an optional staff URL (it powers deep links back to your records), and the username and password of an ArchivesSpace account you control, stored encrypted. A dedicated account keeps Acervo's changes legible in your instance's own history. Then pick the repository. A hosted sandbox lets you rehearse sends end to end before connecting your own instance.

The same settings page asks how your institution describes collections, item records or access points. That choice sets the preselected send mode on every batch's Deliver tab, and you can switch modes on any batch before sending. At send time you choose the target collection or series from a browsable tree of your repository.

Re-sends update records in place, identified by external identifiers Acervo stamps on every record it creates. Retrying a send never duplicates. Every send also leaves a provenance trail. A processing event on the record, and a send receipt in Acervo showing what went where, when.

Item records vs access points

Two send modes, two contracts

• Item records · one archival object per document, created under the target you choose, carrying dates, extents, notes, and linked agents
• Access points · enriches one existing record, additively, with your curated agents and subjects. It never creates records and never removes anything already on the record.

Access-point sends respect archival non-repetition. If a parent record already carries an access point, the send to a child skips it, with a receipt saying so. A checkbox overrides this when you want the repetition.

Housed in, accession, digital access

The About-this-batch strip on the Review tab captures three facts once per batch, and delivery binds them

• Housed in · the box and folder the material lives in. A typeahead searches your repository's existing containers; filename inference reads container numbers out of scan names ("B12_F03", "caja12-carpeta3") and asks you to confirm. At send time, existing containers are reused and missing ones are created.
• From accession · the accession the material came from, found by identifier, title, or donor. The collection gains the accession link, and the accession gains a processing event. Without a send, the fact still flows into EAD acquisition info.
• Digital access · whether digital objects should attach to sent records, driven by the institution profile's URL pattern. Opt out per batch if a batch shouldn't carry them.

Digital identifiers

When your scans live in a digital repository (dLOC, Islandora, your own server), set the URL pattern once in the institution profile. The per-document identifier fills the blank. Identifiers are always supplied, never invented by Acervo.

Two ways in. Type an identifier on the document card's chip, or paste a two-column spreadsheet (filename, identifier) and let Acervo match by filename, then title. The spreadsheet works at both timings. Before sending (digital objects ride the item send) or after (Acervo finds the already-sent records by their external identifiers and attaches digital objects in place).

Where to go next

• §8 Exporting Your Results · The file formats, the configurable ZIP, and the dLOC handoff
• §6 Curating the Batch Roll-up · The curation that everything on Deliver reflects