10d after the breakout, github_stars is 8700 vs 5628 at breakout — still above baseline
What changed
Every entry below is a change that has already been observed, with the numbers that triggered it. Kanon detects change; it does not forecast it.
14d after the breakout, github_stars is 12062 vs 10879 at breakout — still above baseline
14d after the breakout, producthunt_votes is 516 vs 367 at breakout — still above baseline
15d after the breakout, hn_points is 146 vs 64 at breakout — still above baseline
17d after the breakout, github_stars is 3688 vs 879 at breakout — still above baseline
15d after the breakout, github_stars is 14745 vs 14398 at breakout — still above baseline
17d after the breakout, producthunt_votes is 201 vs 0 at breakout — still above baseline
9d after the breakout, github_stars is 8700 vs 5628 at breakout — still above baseline
13d after the breakout, github_stars is 12062 vs 10879 at breakout — still above baseline
13d after the breakout, producthunt_votes is 516 vs 367 at breakout — still above baseline
14d after the breakout, hn_points is 146 vs 64 at breakout — still above baseline
16d after the breakout, github_stars is 3631 vs 879 at breakout — still above baseline
14d after the breakout, github_stars is 14745 vs 14398 at breakout — still above baseline
16d after the breakout, producthunt_votes is 201 vs 0 at breakout — still above baseline
8d after the breakout, github_stars is 8404 vs 5628 at breakout — still above baseline
12d after the breakout, github_stars is 12062 vs 10879 at breakout — still above baseline
12d after the breakout, producthunt_votes is 516 vs 367 at breakout — still above baseline
13d after the breakout, hn_points is 146 vs 64 at breakout — still above baseline
15d after the breakout, github_stars is 3576 vs 879 at breakout — still above baseline
13d after the breakout, github_stars is 14745 vs 14398 at breakout — still above baseline
15d after the breakout, producthunt_votes is 201 vs 0 at breakout — still above baseline
Detector
signal-v1. Published thresholds: accelerating = this window's gain ≥ 1.5× the previous window (min +5); cooling = momentum falls to ≤ 40% of the prior window; reactivated = flat for 7d then moving; anomaly = |z| ≥ 3 against the entity's own history. Every number above is recomputable from these constants. Machine feed: /api/v2/changes.