Skip to main content

Centralized SSO Token — Status 2026-07-19

Headline​

Initiative chunked into two phases to resolve the DRI's Q3 FE over-capacity. Syafrizal M. was over at 1.21 (63 committed ÷ 52 effective, ~11 days over), driven overwhelmingly by this initiative's full FE 23 md. Decision 0002 splits it so only the Launchpad pilot runs in Q3; the three step-5 consumer surfaces move to 2026-Q4. No scope is cut — the total stays 37.5 md.

What changed​

PhaseWindowSurfacesEffort (FE/BE/QA)Total
Phase 12026-Q3Launchpad (pilot, cross-repo anchor)FE 6.5 · BE 0.5 · QA 2.59.5 md
Phase 22026-Q4CRM · Hub · Hub Chat v2FE 16.5 · BE 4.0 · QA 7.528 md

Phase 2 is Q4-deferred and excluded from the 2026-Q3 capacity plan (same mechanism as an as-built exclusion). Capacity effect on the three affected implementors:

  • Syafrizal M. (FE): 63 → 46.5 ÷ 52 = 0.89 (under) — over-commit cleared.
  • Satya (BE): 33.5 → 29.5 (0.57).
  • Yoddi (QA): 49.5 → 42 (0.81).

Still at-risk​

Health stays at-risk — chunking removed the capacity driver, not the external one. Phase 1 (Launchpad) is still blocked on the Session Manager staging endpoint (SB-2) and the per-repo current-company contract, both owned by SSO/Account (BE). No Bifrost task builds these; they gate the pilot enable. The Q3 E2E-fit pressure, however, largely dissolves: only Launchpad's 2.5 md E2E now sits in Q3 (was the full 10 md).

Next​

  • Phase 1 (Launchpad) proceeds as soon as SB-2 lands; it is P12 on the roadmap and FE-queue-late, but now fits Q3 capacity.
  • Phase 2 (28 md) is deferred, not dropped — tracked in the README ## Phasing section and the timeline's Q4 milestones. It gets a fresh forecast + staffing in Q4 planning (no Q4 capacity plan exists yet).

Propagation​

Decision 0002 · README ## Phasing · timeline milestone table (P1/P2 split) · capacity.md (Syafrizal/Satya/Yoddi rows + resolution bullet) · roadmap P12 note + capacity summary.