Peptide Sentinel

Data-led reporting on the peptide research record

Dataset

Federal Register, four compounding terms

Every document the Federal Register API returned for four search terms used on the drug compounding beat, one row per term per document, with the permanent link to each.

673 rows10 columnsGenerated 13 October 2026

The file

673 rows, 10 columns, 286 KB. Comma separated, UTF-8, one header row, quoted where a value contains a comma or a quote. Download federal-register-compounding.csv.

How it was derived

Four live requests to the Federal Register documents endpoint, one per term, ordered newest first and paged to exhaustion. Fields are taken directly from the API response. A document matching more than one term appears once per term, which is why the row count exceeds the number of distinct documents.

Regenerate it with node scripts/sentinel-analysis.mjs --write. The same run produces every figure in 673 hits, 431 documents, so the page and the file cannot drift apart.

Columns

Every column in federal-register-compounding.csv
ColumnWhat it holds
termWhich of the four searches returned this row. Quoted values are phrase searches.
document_numberThe Federal Register document number. Unique per document; the key to de-duplicate on.
publication_dateDate of publication in the Register, ISO format.
typeRule, Proposed Rule, Notice, Correction or Uncategorized Document, as the API reports it.
agenciesAll agencies listed on the document, semicolon separated.
fda_is_an_agencyyes when the Food and Drug Administration appears in that list. A crude relevance proxy, reported alongside the raw count rather than substituted for it.
citationFederal Register citation, where the API supplies one.
actionThe document's own one-line action statement, where it carries one.
titleThe document title, verbatim.
html_urlPermanent link to the document on federalregister.gov.

What it cannot be used for

The term search matches the full text of a document. A hit is a mention, not a subject, and the counts are a snapshot that will not reproduce exactly on a later date.

The limits of the analysis built on this file are stated on the investigation itself, and the rules this publication applies to all of its counting are on the methods page.