National Archives Discovery

The National Archives Discovery

UK government's official archive catalogue with 32+ million record descriptions spanning 1,000+ years. No auth required.

Data Model

• Records — Catalogue entries with id (e.g. C7394009), reference (e.g. GFM 33/2280/5273), title, description, coveringDates, department, heldBy, and closure status.
• Record Details — Extended metadata including scopeContent, citableReference, legalStatus, creatorName, places, corporateNames, and physical description.
• Repositories — Archives, libraries, and museums with id (e.g. A13532670), archonReference, contact details, opening hours, and access information.

Search Parameters

• All search parameters use the sps. prefix (e.g. sps.searchQuery, sps.resultsPageSize).
• Pagination uses sps.page (1-indexed, max 1000) and sps.resultsPageSize (max 100).
• Date filters: sps.dateFrom and sps.dateTo in YYYY-MM-DD format.
• Department filter: sps.departments for narrowing to specific archive departments (e.g. WO for War Office).
• Catalogue levels: 1=Lettercode, 2=Division, 3=Series, 6=Sub-sub-series, 7=Piece, 8=Item.

Gotchas

• Record IDs vs references: Use the id field (e.g. C7394009) for API lookups, not the human-readable reference.
• Repository IDs: Use the id field (e.g. A13532670) for repository detail lookups.
• Closure status: O = Open, C = Closed, R = Retained. Closed records have restricted access.
• Search response facets: Search results include taxonomy, time period, and department facets for narrowing queries.
• Prod egress: The Discovery API returns HTTP 403 HTML ("Restricted") to AWS/datacenter IPs. MCPBundles production routes through the Cloudflare egress relay; local dev works direct unless you hit the same block.