In the trough
Everything the crawler found that isn't listed yet, and why. Paid jumpers are one first-in-first-out line, drained before the free line, which is also first-in-first-out. The Cap'm's own trawled finds wait in a lane of their own, scanned on our OpenRouter bill so they never take a free slot or a neuron from a repo that came to us; commit the file and a trawled repo moves to the free line. Nothing here is votable; everything is readable and reportable. Rejected repos re-enter detection when the owner presses Refresh or anyone pings them after a fix. Missing a repo that has the file? Log in and request a scan at /scan; it tells you why.
Scans are rationed daily so the trough stays free to enter; the count resets at 00:00 UTC. The line is first come, first served, jumpers excepted. Every number behind it is public on /stats.
A new repo needs the sweep to find it, then a scan tick to inspect it. The sweep only sees what GitHub's code search has indexed, which can trail a push. Request a scan to skip the wait.