How the liveness probe pipeline works

Two probes run against every coin in this archive, and it matters to be precise about what each one actually checks, because neither checks whether a coin is still alive.

The first is the GitHub search step described in the companion note on scoring. It looks for a repository that plausibly belongs to the coin and ranks how confident that match is. It runs coin by coin, waiting between requests to stay under GitHub's unauthenticated rate limit, and a run that gets rate-limited partway through can resume later without re-checking coins that already produced a usable result. Nothing from this step is published as fact by itself. It only produces a candidate for the second-stage check described in the other note.

The second probe checks the coin's own legacy subdomain, the one its original block explorer ran on. Every one of the 37 now resolves to this domain's reclaim, either as a direct redirect to the coin's new page or as a fetch that no longer reaches anything at all. That result dates when the reclaim took hold. It says nothing about the coin, because the subdomain belongs to us now regardless of what happened to the project behind it.

What neither probe does, and what nothing on this site does yet, is reach a running node, read a block, or check whether an exchange still lists the coin. That would be a third kind of check entirely, one that asks about the chain directly instead of a repository or a hostname, and it has not been built. Until it is, confirmed on this archive means a source repository was found and its own text supports the match. It does not mean the chain is running, and it does not mean the chain is dead. Every confirmed coin page says so directly, because the gap between what a repository can tell you and what a chain is doing is why this pipeline records dated observations instead of conclusions.