Select a fictional repository, change the observation window, and inspect why its
ForgeRank score exists. Nothing on this page describes a real GitHub project.
Cohort shown Reasons visible Gaps preserved
Observed developer tools
4 fictional repositories · score model v1
Rank #01 / TypeScript
demo-labs/relay-ui
A fictional component library with a growing contributor cohort.
71ForgeRank
high confidence
Observed reach2.5Ksample stars
7-day growth+1.6%3 retained snapshots
Activity38sample commits / 30d
Evidence timeNowsynthetic run
Score reasons / v1
Six weighted dimensions
× 0.92 confidence
Retained observations
Movement, not interpolation
+4.0 pts
OlderCurrent sample
Evidence ledger
What produced this view
demo-sample-v1
How to read ForgeRank
Three checks before you trust a rank.
ForgeRank is designed to expose the boundaries that ordinary leaderboards hide.
01
Check the cohort.
“#1” means first among the repositories included in the named observed set—not first
on GitHub.
02
Check the clock.
Totals and trends can have different observation times. Stale evidence remains visibly
stale.
03
Open the reasons.
The score version, raw inputs, missing dimensions, and confidence multiplier remain
inspectable.
Run the actual application
Use the same isolated sample locally.
pnpm demo prepares a dedicated PGlite database and opens the full product
on 127.0.0.1:3001.
