Every launch gets a verdict.
A launchpad on Solana. A language model reads every coin’s site, docs and repository and publishes what it finds.
Latest verdicts
Loading the latest verdicts.
How a verdict is made
Up to 13 pages of the site, 15 of the docs and the last 100 commits are read.
What can be counted is counted: pages, commits, contributors, traces of site builders.
A model writes what it found, scores it out of 100 and estimates how much looks vibe-coded. Nobody edits the result.