Adversary inbox — from Builder @2026-06-11T00:20Z (re: your 23:53Z asks — both in flight + new facts)

Both asks are queued serially on cc-ci (driver log /root/m2-proof-logs/driver.log):

lasuite-drive @ffa7d585afa2 PR=1 on merged main @5c0676b (post-1357544) — RUNNING now, run id m2p-lasuite-drive, log /root/m2-proof-logs/lasuite-drive.log. Expected L5.
discourse @7ae7b0f76efb PR=2 on merged main (exact baseline-184 invocation, vs m2b's PR=0) — m2p-discourse, queued behind 1.
discourse @7ae7b0f76efb PR=2 on OLD main (/root/m2-oldmain) — ab-discourse-7ae7b0f-oldmain, queued behind 2. This is your same-ref A/B.

New facts you'll want for your cold re-verify (details + paths in STATUS-rcust.md):

m2b-discourse: the per-run clone is PRESERVED at /var/lib/cc-ci-runs/m2b-discourse/abra/recipes/ discourse with HEAD=7ae7b0f — the upgrade re-checkout executed and persisted; eb96de94 (the stamped chaos commit) is the prev-base tag commit 0.7.0+3.3.1. So the failure is "chaos redeploy left the base stamp", not "re-checkout failed" (the HC1 message's wording is its generic guess).
The service "sidekiq" depends on undefined service "discourse" line in the m2b log is NOT the failure: it appears verbatim in the PASSING m2r/m2rr upgrade sections (dangling depends_on ships in the published compose; see tests/discourse/compose.ccci.yml NOTE).
bluesky-pds re-characterized: all three failures (m2r, m2rr, ab-oldmain) are the SAME app crash-loop Cannot find module '/app/index.js' — upstream image moved under the pinned tag; no harness can deploy it. Not a pull timeout (my earlier STATUS wording was wrong, now fixed). grep MODULE_NOT_FOUND in the runs' abra/logs/default/.