GitHub Octoverse and GitHub REST/GraphQL API

Two views of one platform, stacked. The first is Octoverse, GitHub's annual state-of-open-source report - the statistic everyone else quotes. Its most recent edition counts more than 180 million developer accounts, roughly one new developer joining every second, 36 million of them added in a single year, and TypeScript sitting at number one among programming languages for the first time, ahead of both Python and JavaScript.

The second view is the repository graph underneath the report: per-project records carrying stars, forks, open-issue volume, primary language and lifecycle timestamps, with topic, release and dependency context reachable alongside. One layer gives you the sentence for the board deck; the other gives you the rows that survive due diligence.

Datadory delivers both as one systems software dataset - report-era aggregates joined to record-level evidence. Get a sample of this dataset and find out which half your question actually needed.

Claims about GitHub deserve a receipt, so here is one. Queried for operating-system projects, the catalog opens with this record:

full_name          : HeyPuter/puter
stargazers_count   : 43136
forks_count        : 4013
open_issues_count  : 18
language           : TypeScript
created_at         : 2024-03-03T01:44:32Z
pushed_at          : 2026-08-21T06:52:01Z

Read it like an analyst rather than a fan. Forty-three thousand stars accumulated over roughly nine hundred days is about forty-eight a day, sustained, for a project created in March 2024. Four thousand forks means thousands of parties wanted their own copy of the tree. Eighteen open issues against that traffic says triage is keeping pace. And a pushed_at stamp landing on August 21, 2026 - the day this record was captured - marks a repository people are still working in, not a museum piece.

Seven verified fields, one flat record per repository, definitions checked against a live capture during cataloging. Together they cover identity, popularity, classification and lifecycle - enough to score project health without leaving the row.

Geography - global. Octoverse reports developer participation by country each year, so geography arrives as an annual aggregate; repository records themselves are not partitioned by country.

Temporal - three clocks run at once. Annual report editions reach back to 2014, a decade-plus of citable aggregates. Repository records self-describe with creation and last-push timestamps, so cohort windows reconstruct from the rows without an external calendar. And the public event-timeline layer reaches back to February 2011, older than the report series itself.

Granularity - per-repository, per-user and per-event at the record level; per-language, per-topic and per-country at the report level. Most analyses meet in the middle: take a slice of repositories, aggregate it, then benchmark the result against the official figure.

API, files, or your warehouse. Daily, weekly, or hourly.

• Systems-software landscape mapping - count active projects by topic across operating systems, kernels, containers and databases, and watch which categories thicken before analyst reports notice.
• Platform-bet sizing - 180 million-plus accounts and a language-share leader that just changed for the first time quantify where developer attention is moving while it is still cheap to follow.
• Dependency due diligence - the stars-to-open-issues ratio, fork depth and push recency in one row screen a candidate library in seconds instead of a sprint.
• Developer-mindshare tracking - language and topic growth curves give tooling and infrastructure vendors a bottom-up demand signal years ahead of revenue data.
• Citation-grade research - a decade of annual editions plus event history to February 2011 yields reproducible series that survive peer review.

Data scientists and ML engineers get event-grain history deep enough for time-series work and a flat record shape that trains on without cleanup theater. Competitive-intel and product teams read rivals' ecosystems as fork and issue velocity, which tends to move before launch announcements do. Investors and quants treat language-share shifts - TypeScript taking the top slot is the freshest example - as a leading indicator of where developer labor is reallocating. Market researchers get the statistic everyone quotes, 180 million developers joining one per second, backed by rows they can audit. Developers and builders wire the record shape straight into dependency checkers and ecosystem dashboards.

Systems Software catalog neighbors that answer adjacent questions. PyPI - Python Package Index API and Stats counts package versions and downloads rather than repositories, and the head-to-head comparison is worth reading before choosing a unit of analysis. GitHut - GitHub Language Statistics renders the language-share picture in aggregate form. npm Public Registry API covers the JavaScript package economy that TypeScript's rise feeds. deps.dev - Open Source Insights API scores transitive dependency health once you have picked candidates.