National Archives Catalog MCP for AI. Search Millions of Historical Records & Contributions

WHO IT'S FOR: The National Archives Catalog MCP Server connects your AI agent directly to the US National Archives (NARA) collection. You can search through millions of historical documents, photos, and records using natural language. It lets you retrieve metadata for specific IDs, track contributions made by others, navigate record relationships, and even manage community additions like creating a new tag or transcription.

SEARCHING AND DISCOVERING RECORDS:
You can find records across NARA using general keywords or URL parameters with search_records. If you're looking for something specific in the main body text, use search_records_by_text. To narrow it down to full transcriptions, run a search using search_records_by_transcription. You can also target records by community input: find items linked to a specific tag with search_records_by_tag, or limit your scope only to those records that mention a certain comment ID with search_records_by_comment.

For the deepest dive, you can use search_contributions to query across all user input—tags, comments, and everything in between.

MANAGING RELATIONSHIPS AND METADATA:
You'll know exactly what belongs together. Use get_record_children to pull a list of immediate records that are related to a specific parent ID (NaId). You can pull usage statistics on contributions for any record using get_record_stats. To track down the details on one single tag, run get_tag, or see every tag currently applied to a record by calling get_tags_by_naid.

You'll also get all tags assigned by a specific contributor user ID using get_tags_by_userid.

ACCESSING COMMUNITY CONTRIBUTIONS:
The system lets you track down every comment associated with one NaId via get_comments_by_naid, or check out the full history of comments made by a particular contributor using get_comments_by_userid. You can retrieve all contributions—including tags and comments—for both a specific record (get_contributions_by_target_naid) and a specific user (get_contributions_by_userid). For full transparency, you've got access to get_justifications, which retrieves the official reasons for any administrative action taken on records.

You can also check what recent notifications are sitting in a user’s inbox using get_user_notifications.

VIEWING TRANSCIPTIONS AND USER DATA:
You'll get all transcriptions related to one record with get_transcriptions_by_naid, or see every transcription created by a specific contributor using get_transcriptions_by_userid. You can fetch the full text of a document using get_transcription and review its change log history via get_transcription_history. For user profiles, you can grab public data for one person with get_user, or pull details on multiple users at once using get_users, which filters by role or email.

MODIFYING AND ADDING CONTENT:
Want to contribute? You can add a brand new public comment to any record ID using create_comment. If you need to tweak that text later, use update_comment. To manage tags, you'll either create_tag or remove one using delete_tag. Similarly, if you find a document that needs transcribing, you can add the full text with create_transcription, and then modify it later with update_transcription.

You also have tools to clean up contributions by deactivating multiple records inputs at once using delete_contributions.

CLEANING UP AND DELETING:
To remove a comment, you run delete_comment. If you need to take down a tag that was attached incorrectly, use delete_tag. You can also pull the list of all current announcements posted to the catalog using get_announcements or get_latest_announcements, and check out the general status metrics for any given NaId with get_record_stats.