FUTURE (recommend, separate from skew fast-follow): enable next.config deploymentId + Vercel Skew Protection (if Pluto's plan allows) for a SOUND client-vs-server build compare — gives the app a real origin-build signal (today the build-ID is COARSE: app sees only its own serving build, never the client's origin build, which is why an app-side skew downgrade is unsound and the daemon owns the deploy-timing discriminator). Also makes a real end-to-end deploy-skew test possible. Low pri.

Sub-tickets

Questions

Activity

• wi-cli-venus created · 2026-06-21
• wi-cli-venus commented · 2026-06-21
  ELEVATED future->PREREQUISITE (3-PM brainstorm, audit-venus's 3-signal predicate correction): the SOUND skew-downgrade predicate is 3-signal = RSC-soft-nav AND manifest-match AND build-id MISMATCH. The 3rd AND is what distinguishes a real data-miss on a valid DYNAMIC route (e.g. /administrar/[id] dead id → pattern matches + RSC, but genuine notFound()) from actual skew. Pluto has NO build-id signal wired today (app sees only its own serving build). So WITHOUT deploymentId, any app-side downgrade silences real dynamic-route data-misses (EVO-12 hole); the daemon deploy-TIMING window is a LOSSY proxy that silences real data-misses within ~10min of a deploy. deploymentId+Skew-Protection is therefore the prerequisite for a SOUND skew-downgrade, not optional. Still gated behind Q3-enable; bump priority when the skew fast-follow (PLUTO-169) is actually scheduled to flip ON.
• wi-cli-venus commented · 2026-06-21
  REFRAMED: MOOT on Hobby. Verified Pluto=Hobby (pluto+mars share one Vercel team team_WG6Gxa6.., per-team billing, mars=Hobby). Per mars-audit circularity (ratified pm-llmmsgsrv v2): x-deployment-id build-id header is Vercel Skew-Protection=Pro+ only, AND circular (Skew Protection when ON version-pins stale clients so the skew-404 doesn't fire). So deploymentId+Skew-Protection is build-id-present-where-moot / absent-where-needed. Pluto's skew signal-3 = build-AGE timer (daemon-side, durable canonical NOT fallback). PLUTO-170 only relevant IF Elazar ever upgrades to Pro, and even then build-id is belt-and-braces not load-bearing. Not a prerequisite for the skew-downgrade. Deprioritized.