Skip to main content

CDP Sprint 2 — Jira reconciliation pass-04 (2026-07-27)

Board: sprint 21484 (CDP 26Q3 Sprint 2) · Window: 2026-07-21 → 2026-08-04 · day 7 of 14. Engine: qontak-docs @ 74046ff (past both 30d8080 discipline-resolution and 7bfdc11). Worklist: 1 deterministic / 11 editorial findings over 73 issues.

Flow vitals: 73 issues (72 → 73 since pass-03), WIP 22 (up from 16), done-since throughput 0 (since 2026-07-26). Oldest in-progress: TF-3283 / TF-3289 (Giovanni Massimiliano / Jovi, Export FE, 33d) then TF-3271 (Rizki Alfi Ramdhani / Alfi, Team Owner Mobile, 33d). Zero completions in the last day, heading directly into the P1 Segmentation 07-28 hard commitment (tomorrow) with an unresolved burnup question (see below) — this is the most urgent item this pass surfaces.

Process note on this pass's predecessor. Pass-03 (2026-07-26, compacted run) lumped F-1 and F-3 into a generic "carryover — STILL OPEN" line without re-verifying either against live Jira. Both were correctly CLOSED with hard evidence in pass-01 (2026-07-23) and remain resolved on independent re-check this pass (below). This pass corrects that regression; it is not new drift. Flagging so the same shortcut isn't repeated: a "carryover" label is not a substitute for re-testing each id per the skill's carry-forward rule.

Jira reconciliation — recheck (2026-07-27, pass 4)​

Status of prior findings (by id):

  • F-1 (coverage-gap) — RESOLVED (re-confirmed; pass-03's reopening was itself in error — see process note above). TF-3466–3472 (Legacy Migration Activity Logs Tasks 3–9) are all live in sprint 21484 under Epic TF-3182 (To Do). The board-scoping gap stays closed.
  • F-3 (data-quality) — RESOLVED (re-confirmed). TF-3459 verified via direct Jira lookup: status Won't Fix, resolution Done. Still correctly 0-credited, nothing further blocks.
  • F-4 (coverage-gap) — ESCALATED — 9th consecutive pass, zero board movement. Epic TF-3302 (CEBE) still has 0 issues in sprint 21484; searching parent = TF-3302 across all Jira returns only unassigned, un-sprinted auto-synced Stories (TF-3354–3358), never an RFC-authoring Task. ADR 0001 exists (authored 2026-07-23) but is still status: proposal — no decision recorded. Per the ADR's own text, a ninth pass with zero movement is Option B (deprioritize for 26Q3) by default. Owner: Julio Jeffer / PM (Zhelia) — this pass treats the default as now in effect unless the ADR is resolved to Option A before sprint close.
  • F-5 (status-progress, residual) — STILL OPEN, unchanged. The task-breakdown doc explicitly declares itself Backend only (label: backend under TF-3185). FE (chunks 8–9), Mobile (chunk 10) ... are out of scope for this doc — the FE/Mobile chunk mapping Ghozi/DRI owe has not been added. Owner: Ghozi / DRI, unchanged.
  • F-7 (status-progress) — ESCALATED, unchanged (4th pass). Dated notes-mention-user/delivery/status/2026-07-26.md explicitly restates: "TF-3506 still awaits a retro-cost or out-of-band decision." Owner: DRI / PM, unchanged.
  • F-8 (scope-watch) — ESCALATED, unchanged (4th pass) — now time-critical. TF-3507/3508/ 3511 (datamart hardening) remain Done with no in-scope/retro-cost decision recorded. Separately (same root commitment): TF-3401 (the FE item the plan's blocker §1 still calls "Blocked") has actually been Done since before pass-01 (2026-07-23) — the dated customer-segmentation/delivery/status/2026-07-26.md already flags that the carried 4 days / 47% baseline is BE-only and excludes the now-done FE tasks (TF-3399/3400/3401), so the true progress is under-counted, not fully unknown — but the DRI recompute pass-01 routed 4 days ago still has not landed. With the 07-28 hard commitment one day away, this recompute (and the F-8 retro-cost call it's entangled with) needs the DRI today, not at sprint close. Owner: DRI (Zhelia).
  • F-10 (off-plan-load) — RESOLVED (carried from pass-03). roster.json / capacity.md carry Wulan and Zhelia.
  • F-14 (epic-wiring-drift) — STILL OPEN — partial movement. TF-3550 now carries an Epic (TF-3500, the engineering bucket) where before it had none — but TF-3500 is not the correct initiative Epic for an Export Customer Data layout bug; it should be TF-3181. Owner: DRI (Zhelia) — reparent TF-3550 to TF-3181, or confirm it's deliberately bucket-scoped work outside Export.
  • F-15 / F-16 (coverage-gap / assignment-drift) — RESOLVED. This pass's deterministic recompute emitted zero coverage-gap / assignment-drift rows for cdp/adhoc-q3-2026 or cdp/engineering-q3-2026 — the generic "still lags" note in pass-03 carried no cited evidence; this pass's clean recompute is the stronger signal. (Separately, see new F-29 below — a distinct staleness in the sprint plan's per-engineer queue, not the initiative trackers this finding covers.)
  • F-17 / F-19 (assignment-staffing / data-quality) — RESOLVED (carried from pass-03).
  • F-20 (off-plan-load) — RESOLVED. roster.json's "Wulan Febyazzahra Putri (Bulan)" key now exact-matches the live Jira assignee name; no off-plan-load fired this pass.
  • F-21 (roster-mismatch) — RESOLVED — with a caveat worth recording. The engine still emits "Izzul: 0 issues assigned" this pass, but live Jira shows Izzul actually owns TF-3570 and TF-3571, both Done. The "0" reading is a structural artifact: both issues are parented to Epic TF-3521 (QA BAU), which sits outside the tracked-initiative map the engine's staffing join walks — the same root cause as F-28 below, not a real staffing gap. Izzul is not idle; treat this as resolved via the same TF-3521 disposition, and expect the engine to keep surfacing it every pass until TF-3521 is registered as a tracked (if off-roadmap) initiative.
  • F-22 (roster-mismatch) — RESOLVED. Wulan's TF-3550 now counts under bucket Epic TF-3500, which the engine's staffing join does track; the finding no longer fires.

New findings from the 2026-07-27 pull:

  • F-23 / F-24 / F-25 / F-26 / F-27 (code-evidence-gap, engine-emitted): five board-done issues have no code trace in any of the six scoped repos (contact-service, hub-chat, qontak-customer-fe, qontak.com, mobile-qontak-crm, mobile-qontak-chat).
    • F-23 TF-3568 (Won't Fix) and F-27 TF-3548 (Won't Fix) are genuinely code-free by resolution — a Won't Fix ticket was never built. TF-3548 is already covered by F-5's disposition (blocked on CRM→USMAN migration). No action beyond that.
    • F-24 TF-3555, F-25 TF-3572, F-26 TF-3584 are Done (not Won't Fix) with zero code trace — these need a TPM/assignee confirmation of "genuinely code-free (e.g. a verification/ops task)" vs "stale board status." Owner: TPM, route to the respective assignees (Azani for TF-3555, unclear owner for TF-3572/TF-3584 — confirm via Jira).
  • F-28 (epic-wiring-drift) — 5 sprint issues (TF-3578, TF-3579, TF-3580, TF-3570, TF-3571) parented to untracked Epic TF-3521. Formalizing into the ledger as RESOLVED (documented off-map exception): the sprint doc's own "Epic-wiring disposition (2026-07-26 pass-03)" already declares TF-3521 the intentional executable container for CDP QA BAU. No further action — this id exists so the recurring engine row doesn't read as fresh drift each pass.
  • F-29 (plan staleness, new) — Azani's sprint-plan queue row #3 ("CDP 26Q3 Engineering Initiatives — FE") states "No FE-scoped items in the bucket today ... FE column is —." Live Jira and the bucket tracker (engineering-q3-2026/README.md A11/A12/A13, added 2026-07-22) both show Azani already owns three FE-scoped bucket items: TF-3555 (Done), TF-3556 and TF-3557 (Ready to Deploy). The sprint plan (dated 2026-07-21) predates these by a day and was never corrected in the 2026-07-26 pass-03 update either. Owner: sprint-planner (amend Azani's row 3 to reflect real, already-in-flight bucket work) — this is plan-structure drift, not a Jira problem, so it does not route through Jira remediation.

Closing line — what this pass does not fix. No repo file is edited by this pass file itself; every item above is either already resolved with cited evidence, or named with an owner and a durable home (the CEBE ADR for F-4, the dated status files for F-5/F-7/F-8, TF-3550 reparenting for F-14, the code-evidence confirmations for F-24–26, and a sprint-planner amendment for F-29). The sprint doc's ## Jira reconciliation — findings ledger is rewritten in place from this pass.

Downstream propagation worklist (this pass)​

#FindingCanonical artifact to reconcileOwning tool / skill
1F-4customer-event-based-engine-cebe/delivery/decisions/0001-...mdDRI/PM decision — remap ADR status:
2F-5team-owner-field-and-team-permission/rfcs/*.task-breakdown.mdGhozi/DRI — add FE/Mobile chunk mapping
3F-7notes-mention-user/delivery/status/<date>.mdDRI/PM — retro-cost or out-of-band decision
4F-8customer-segmentation/delivery/status/<date>.mdDRI — retro-cost decision + burnup recompute (time-critical, 07-28 commitment)
5F-14Jira: TF-3550 parent fieldDRI — reparent to TF-3181 (board fix, not a repo edit)
6F-24 / F-25 / F-26Jira: TF-3555 / TF-3572 / TF-3584TPM + assignees — confirm code-free vs stale board
7F-29delivery/sprints/2026-07-21-sprint-02.md Azani row 3sprint-planner amendment

Run npm run lint:delivery after any of the above artifact edits land.