venus go-live: place ~/.config/bwi-client.env (venus write-key, delivered out-of-band) so venus agents route wi/wi-web → wi-http:8042. Until then venus hard-fails safe (no stub) per #645.

Sub-tickets

Questions

Activity

• wi-cli-whey created · 2026-06-05
• wi-cli-whey statusChanged · 2026-06-05
  Driving now: pm-venus-cc surfaced a split-brain (writing a stale/local store, venus mis-mapped to project 7). Need: provision venus write-key + place ~/.config/bwi-client.env on venus → whey wi-http. Asked Elazar to greenlight key provisioning.
• wi-cli-whey spec · 2026-06-05
  Confirmed split-brain (pm-venus-cc): venus host wi runs FILE-mode vs a stale local basquetWorkItems.sqlite (venus=local id7, was empty pre-today); local email ticket #336 canceled, never reached whey, no canonical pollution. venus holding all local writes until wired. Go-live steps: (1) git pull latest on venus (a861e01: wi self-sources env + refuses stub); (2) place ~/.config/bwi-client.env with venus write-key (provision via bin-whey-cc, deliver out-of-band); (3) retire venus's local sqlite so FILE-mode can't fall back. Then venus=project8 on whey, re-file email ticket → VENUS-2. BLOCKED on: Elazar greenlight to provision venus write-key + agent to place env on venus.
• wi-cli-whey completed · 2026-06-05
  venus go-live complete + verified: provisioned wi-cli-venus write-key (/etc/bwi-http.env, wi-http restarted), placed ~/.config/bwi-client.env on venus (chmod 600), git pulled venus→a861e01 (wi self-sources env), trashed venus local sqlite. Verified: venus wi self-sources + hits whey (venus=VENUS, 12 items); email ticket re-filed as VENUS-2. Residual fleet cleanup (tracked .RETIRED git rm + SQLite hard-error guard) → coder, separate WI.