Numbers computed from the daily snapshots the pipeline keeps, for the same filter and period as the list. A metric the pipeline has not recorded shows as an em dash rather than a zero: “not measured” and “measured as nothing” are different claims.
See the repository list for every tracked repository, ranked by the stars it gained.
Ranked last year
7,097
repositories with a computed delta
Stars in view
74.5M
current totals, not a period
Stars gained last year
+17.6M
7,095 gaining · 1 losing · 1 flat
Pushed in the last 24h
22%
13,622 of 63,099 tracked
Of the +17.6M stars gained across 7,095 repositories last year, the top 10 took 9% and the top 100 took 30%. Repositories that lost or held stars are not part of this total.
A net star change is not proof that users moved between projects: we sample star counts once a day and never record who unstarred what.
+17.6M stars gained across 7,095 repositories last year.
Repositories that lost or held their star count are excluded from the total above, so the shares describe the attention that arrived, not the size of the tracked set.
Which languages gained the most stars last year, and how much per repository. Per-repository is the honest comparison: many small projects can out-gain a few large ones without any project in it being hotter.
Current state of every tracked repository, not the period above: this is what the pipeline saw at the last snapshot.
24h: 13,622 · 7d: 28,850 · 30d: 15,680 · older: 4,947
The shape of the tracked set matters when reading any average above: most repositories are small, so an average star gain is pulled by a handful of large ones.
Share of a repository's stars whose owners also watch it: a watch asks for notifications, a star is closer to a bookmark. Measured on the 55,668 tracked repositories with 100+ stars that have a watcher count; repositories the pipeline has not measured are left out rather than counted as zero.
Only one capture of this trending period has been stored so far, so there is nothing to compare yet.
Week-by-week lines for topics and collections, plus the groups that used to be the story.
Group trends looks backwards over a year of weekly star history, not at this snapshot.
Every tracked repository, from today's snapshot: the period tabs above scope the other panels, not this one.
| Tag | Repos | Today | Per repo |
|---|---|---|---|
| 566 | 2448 | 4.3 | |
| 566 | 2408 | 4.3 | |
| 694 | 1860 | 2.7 | |
| 351 | 1650 | 4.7 | |
| 507 | 1407 | 2.8 | |
| 227 | 1356 | 6 | |
| 335 | 1158 | 3.5 | |
| 104 | 1095 | 10.5 | |
| 774 | 1082 | 1.4 | |
| 64 | 844 | 13.2 | |
| 296 | 775 | 2.6 | |
| 453 | 691 | 1.5 | |
| 259 | 647 | 2.5 | |
| 532 | 640 | 1.2 | |
| 133 | 616 | 4.6 | |
| 774 | 614 | 0.8 | |
| 207 | 594 | 2.9 | |
| 823 | 579 | 0.7 | |
| 1276 | 558 | 0.4 | |
| 80 | 552 | 6.9 | |
| 130 | 495 | 3.8 | |
| 403 | 492 | 1.2 |
Counted from the topics repositories carry today: the pipeline rewrites them every run, so this is not a history. A pair appears only when at least 20 repositories carry each topic; below that the share swings with a single repository.
Counting which topics share repositories across the whole tracked corpus…
