About

OpenIntel Archive is a cross-archive evidence intelligence platform for declassified primary-source documents. It exists to answer one question: what did different intelligence and government organizations know, believe and report about the same event at the same time?

What is in the corpus

Normalized documents from national archives, intelligence agencies and legislative inquiries, each with its archival reference, originating body, date and a direct link to the archive's own copy. On top of the documents sit verbatim passages, typed claim/evidence relations, dated agency assessments, detected divergences between agencies, and information-provenance links that mark when two records repeat a single original source rather than corroborating one another independently.

The rules the evidence is held to

• Quotations are verbatim. Nothing is paraphrased and presented as a quotation.
• A record whose text has not been retrieved is labelled metadata only, and no quotation is attributed to it.
• A passage quoted from secondary scholarship rather than the archival file is labelled as a secondary source.
• Every claim/evidence edge is anchored to a specific passage and carries a basis: documented, inference, or speculation. These are never shown with equal weight.
• Repetition is not corroboration. Where records descend from one original source, the provenance layer says so.

Cite the document, not us

Each document page carries the archival reference and a link to the originating archive. When you quote, attribute the archive. Verify against the original before publishing.