Primary source documents
A dedicated content type for FOIA records, financial filings, court documents, and other primary sources. Documents live in their own collection and index, separate from articles.
Primary source documents are an investigative journalism content type. They are the raw materials that reporting is built on: FOIA responses, financial disclosure filings, court dockets, government records, transcripts. Project Broadsheet gives them their own collection, their own layout, and their own index page at /documents/ so readers can find and cite them independently of the articles that reference them.
Create a document
Documents live at src/documents/:
src/documents/city-housing-fund-audit-2025.md
The directory data file (src/documents/documents.json) sets layout: layouts/document.njk and section: Documents automatically — no need to add those to individual files.
Front matter
---
title: City Housing Fund Audit — FY 2025
description: Inspector General audit of the Springfield housing fund for fiscal year 2025.
docType: financial
source: City of Springfield Office of the Inspector General
date: 2025-11-14
dateObtained: 2025-10-01
fileUrl: /assets/documents/city-housing-fund-audit-fy25.pdf
fileName: OIG-FY25-0042.pdf
relatedArticles:
- /news/housing-fund-audit-findings/
- /analysis/where-the-money-went/
---
Summary or full text of the document in Markdown. For scanned records, summarize key findings here and let `fileUrl` point to the original PDF.
| Field | Required | Notes |
|---|---|---|
title | Yes | Descriptive title for the document |
description | No | One-sentence summary shown in the document header |
docType | Yes | One of financial, legal, foia, transcript, government, correspondence, other |
source | Yes | The issuing body or organization |
date | Yes | Date of publication on your site (used for sorting) |
dateObtained | No | Date the document was received or became public |
fileUrl | No | Path to the original PDF or file in src/assets/documents/ |
fileName | No | Display name for the download link — defaults to "Original document" |
relatedArticles | No | Array of URL paths of articles that reference this document |
The documents index
/documents/ lists all documents by type, with the document type, source, and date visible in the card. Readers can filter by document_type.
Documents do not appear in section feeds, RSS, or the main sitemap. They have their own /documents/feed.xml.
Linking from articles to documents
In article body text, link to the document the same way you link to any page:
According to the [FY 2025 audit](/documents/city-housing-fund-audit-2025/), the fund...
If the document has related_articles set, a "Referenced in" section appears at the bottom of the document page linking back to those articles.
Hosting full PDFs
Put the PDF in src/assets/documents/ and set fileUrl and fileName in front matter. The layout renders a "Download" row in the document header automatically.
fileUrl: /assets/documents/city-housing-fund-audit-fy25.pdf
fileName: OIG-FY25-0042.pdf
The Markdown body acts as a structured, searchable index entry. The PDF is the complete record. This keeps the document discoverable by Pagefind without depending on PDF text extraction.
What to do next
- Writing articles for how to reference documents within article prose.
- Corrections for updating inaccuracies after publication.
- SEO and structured data for how documents appear in search results.
Browse Support for community channels and paid support options, or book a call if you'd like me to set it up for you.