Machine-readable access

The IARCCUM Archive provides two machine-readable services for discovering and retrieving public archive metadata. These services supplement the website, which remains the authoritative publication interface for the archive.

Researchers should use the canonical IARCCUM record URL when citing an archive item.

Model Context Protocol (MCP)

The IARCCUM Model Context Protocol service provides structured, read-only access for compatible artificial intelligence and research tools. It allows these tools to search the archive and retrieve relationships directly from the catalogue rather than reconstructing them from webpages.

MCP endpoint: https://iarccum.org/mcp

The endpoint uses Streamable HTTP. It does not require an IARCCUM account or authentication.

Available MCP tools

  • search_documents: searches public documents by title, description, indexed full text, dates, authors, and organisations.
  • get_document: retrieves the structured metadata for a public document.
  • search_names: searches people, organisations, commissions, and other named bodies.
  • get_name: retrieves a name authority record and its public relationships.
  • search_organisations: searches organisations and archival fonds.
  • find_documents_by_protocol: finds documents using dialogue and archival protocol numbers.
  • get_related_documents: retrieves documents explicitly identified by the archive as related to a specified document.
  • search_events: searches dialogue meetings and other events by title, description, participants, location, and date.

Connecting an MCP client

An MCP client that supports Streamable HTTP can connect directly to the endpoint. A typical configuration is:

{
  "mcpServers": {
    "iarccum-archive": {
      "type": "streamable-http",
      "url": "https://iarccum.org/mcp"
    }
  }
}

The service is registered in the official Model Context Protocol Registry as org.iarccum/archive.

MCP access policy

  • The service is read-only and does not provide arbitrary database access.
  • Only public archive records are returned.
  • Embargoed documents and their restricted contents are not exposed.
  • Results contain canonical IARCCUM URLs so that researchers can consult and cite the normal archive record.
  • Search results are paginated and limited to protect the resources of the website.
  • The endpoint is subject to a basic limit of approximately 60 requests per IP address per minute.

Open Archives Initiative Protocol for Metadata Harvesting

The IARCCUM OAI-PMH endpoint provides Dublin Core metadata for eligible public document records. It is intended for libraries, union catalogues, metadata aggregators, and other scholarly discovery services.

OAI-PMH base URL: https://iarccum.org/oai

The endpoint supports OAI-PMH 2.0 and exposes metadata using the oai_dc metadata format.

Supported OAI-PMH requests

The public document set is identified as documents. Large result sets are divided using OAI-PMH resumption tokens. Harvesters may also use the standard from and until parameters for selective harvesting by datestamp.

Document identifiers follow this pattern: oai:iarccum.org:doc/{ID}

For example, the Dublin Core record for document 263 may be retrieved with: OAI-PMH Dublin Core record for document 263

OAI inclusion policy

The OAI-PMH service includes eligible public document records from the IARCCUM Archive. Inclusion is determined by the document type, the individual record’s OAI setting, and its public access status.

  • Embargoed documents are not included.
  • Records that are not ready for public harvesting may be individually excluded.
  • Some document types, including material for which IARCCUM is not the responsible repository or publisher, are excluded from harvesting.
  • Deleted or withdrawn OAI records are represented according to the endpoint’s persistent deletion policy.
  • Books and serial publications are not harvested as independent IARCCUM records. Their metadata may appear as part of the description of an eligible archive document.

Choosing a service

Use MCP for interactive research, archive searches, protocol-number lookups, name and organisation records, events, and relationships between records.

Use OAI-PMH for systematic metadata harvesting, library discovery systems, catalogue ingestion, and periodic synchronization of public document records.

Use the website to read the complete public record, view available files and publication details, follow related archive resources, and cite the authoritative record.

Canonical record addresses

  • Documents: https://iarccum.org/doc/{ID}
  • Persons and named bodies: https://iarccum.org/name/{ID}
  • Organisations and fonds: https://iarccum.org/org/{ID}
  • Events: https://iarccum.org/event/{ID}

Additional machine-readable guidance

A concise description of the archive for automated research systems is available at llms.txt.

Rights and responsible use

Access to machine-readable metadata does not alter the copyright or use conditions attached to archive materials. Rights statements are included in individual records where they have been established. Researchers should consult the canonical document record before quoting, reproducing, or redistributing an item.

Questions about machine-readable access or archive metadata may be directed to the IARCCUM editor at jesson@ecumenism.net.