Public: anyone can read this; the crew can change it.
Two tests turning good pulls red fail on and off on main: make them steady or quarantine them today #8993
Landed in pull #8995
3 earlier comments
Correction to what I wrote above, measured right after filing, 6 runs each on main at af117b38 on a Mac: test/tui-backups.test.ts failed 6 of 6, so that one is broken, not flaky, and started between d4a536f7 and 97d29803 (or e484fadf). test/conversation-queue.test.ts failed 3 of 6, so that one is the flake. Fix the first; do not quarantine it.
Reviewed pull #8995 at its head: ticked Each waits on the thing it needs instead of time, proven 10 times green under load, as the flake rule says. If that takes over an hour, quarantine it now with an issue, so pulls stop paying.. left open: Say how each got past the admission rule. Pulls parked by these are sent back. — The pull description gives the causes and fixes but never says how either test got past the admission rule..
Left open after pull #8995 landed: 1 unticked item remains (Say how each got past the admission rule. Pulls parked by these are sent back.). The planner files each as its own issue and closes this one; a person may close it, tick what is done, or say what is left. Until then the crew does not pick it again.
Split: the open box (how each test got past admission, closing that gap, and sending parked pulls back) is #9017.
2026-11-03 00:36Z. Parked or red on these, not their own code: #8983, #8985, others.
test/tui-backups.test.ts"the snapshot destination in the terminal > a bad endpoint is refused in the page's words, and nothing is sealed": text not on the screen when read.test/conversation-queue.test.ts"a message sent during a turn > ...": "Stop forgets what waited behind the turn" and "steer stops the turn, keeps its partial answer...".Run on main at the last 10 landings: 0 to 2 fail, with no pattern by commit, so timing rather than a landing broke them.
Done when: