Data downloads

The charts, agency tables and statement pages draw on these files. They are rebuilt with the site; this copy was built 13 Oct 2026.

The files

statements.csv

One row per tracked agency (135 rows): who it is, whether it has published, its dates, the Standard checklist and its current profile. Start here.

statements.json

The 121 published statements in full detail: the current profile (including the text of each commitment), and every version's classification, summary and changed answers.

timeline.json

Every change we recorded (431, newest first), with its classification and summary.

adoption.json

The monthly figures behind the charts on What the statements say, plus each agency's changes to those items and the version they happened in.

Reading the CSV

The columns come in four groups. abbr to originality describe the agency: name, portfolio, whether it must publish (scope), whether it has, statement URL, when we first saw it, how often it has changed, and its originality score. stated_first_published to annual_review_overdue are the statement’s own dates and whether its annual review is overdue. The standard_* columns show which of the Standard’s eight elements the statement covers. The rest are its current profile, using the answer lists in profiles.py, with multiple values separated by semicolons.

An agency with no statement has only the first group filled in, so a blank profile field means there was nothing to read, not that the answer was no.

Before you analyse

The profiles, classifications and summaries are a language model’s reading, so check them against the statement text before you publish a finding. Reading the data explains what each figure means. The record starts in November 2025, so a statement’s first tracked version may not be its first published one.

Each JSON file carries generatedAt, headSha and schemaVersions, so you can tell which version of the pipeline produced it. New CSV columns are added at the end.

Statement text

The text of every version is in the git repository. The current statements are also published on the AT Protocol network (which Bluesky runs on) by @apsaitracker.app, with one record per statement and one per change, each with a permanent identifier. You can browse the records or subscribe to the network’s live feed.

Licence

These files are CC BY 4.0. See About for how to cite the tracker.