bvandeusenandClaude Opus 5
b3e4491b0a
CI and images / lint (push) Failing after 2s
CI and images / extension-version (push) Successful in 3s
CI and images / frontend-build (push) Successful in 20s
CI and images / backend-lint-and-test (push) Successful in 32s
CI and images / integration (push) Failing after 2m18s
CI and images / sign-extension (push) Skipped
CI and images / build-web (push) Skipped
CI and images / smoke-web (push) Skipped
CI and images / promote (push) Skipped
CI and images / build-agent (push) Skipped
The card's copy had gone stale against its own backend in three places, which
matters more than usual here because every line of it is a claim about what FC
will do with the operator's library:
* "nothing is linked until you accept one" is no longer true for a pair the
creator's own file name settles. The switch that governs that is now on the
card rather than only in the database.
* "always needs two reasons — close in time and the post mentioning Discord"
left out the shared-marker signal, and named 0.55 as the number below which
no single signal can carry a pair. That number is 0.45 now.
* "pairs only appear when a post lands near a drop and says it is about
Discord" described the empty state as if the identity route did not exist.
A queued row now shows the working name behind it when there is one, and says
in the same breath why that name was not enough on its own. A review queue
that cannot explain itself is one the operator learns to click through without
reading, and this is the only signal that claims the two posts are the SAME
piece rather than that they happened near each other.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LVjrnpQjRgHdvq95rASoiR