Data, sources & trust engine.
cyrus is a static index with a live conscience: every number below was pulled from a public API, regenerated on a schedule, and can be re-verified by you in one click. Here is exactly what is running, and on what cadence.
Active snapshot verifications
running now · 8.8k archived program projects · 616 organizations named by a program archiveAccepted project records, 519 organizations, 2016-2026
Source: gsocorganizations.dev - a community mirror of Google's per-year org pages
1,303 mentored projects hosted by 47 Linux Foundation projects
Source: api.mentorship.lfx.linuxfoundation.org - every published project
437 internships across 179 communities, plus 66 Bitcoin projects
Source: Outreachy archives, its call for participation, and event organizers
423 projects in 420 repositories, 47,896 contributors on the board
Source: gssoc.girlscript.org/api - projects, mentors and leaderboard
Organizations that took part, plus 306 curated wishlist repositories
Source: 24pullrequests.com projects.json + organisations.json
120 upstream organizations behind MLH's forks, 87 KDE ideas with their mentors
Source: Every MLH fork and the upstream it came from; the KDE wiki ideas pages
12 programs' own records, each row keeping the source it was read from
Source: public/data/program-projects/ - one shard per program, indexed by scripts/build-program-projects.mjs
Six rules the index polices itself with.
Stars first, always shown beside forks, open issues and last push - so a dormant 200k-star repo can't pretend otherwise.
Beginner under 10k stars, intermediate to 50k, advanced above. Labels come from the data, never from a claim in a README.
Every org avatar is served straight from GitHub's own CDN. Program covers are our own generated SVG artwork - no copied branding anywhere.
Demo sign-in keeps saved issues and repositories in your browser's localStorage. With a real session, they sync to one private row keyed by your account - no password, no stored access token, no tracking pixel.
When an API is rate-limited, pages fall back to deterministic estimates and say so in plain words on the card.
Not affiliated with GitHub, Google or Anthropic. If a snapshot mislabels your project, we fix it on the next regeneration.
Sources & cadence
✦4 live feedsThe current snapshot, in plain numbers
Spot something the snapshot got wrong? The About page explains exactly how data is gathered and how to tell us - corrections land in the next regeneration, not in a ticket queue.
The data comes from teams that ship at scale.
Every org profile, star count and program record on cyrus is pulled from the public GitHub footprint of organizations like these - then re-verified on a schedule.
