Skip to main content

Chatbot Sprint 1 (2026-Q3) — Jira reconciliation, pass 2 (2026-07-15)

Jira reconciliation — recheck (2026-07-15)​

Second resync (sprint:reconcile --team chatbot --json, qontak-docs). Sprint 21473 now holds 54 issues (up from 22 at pass 1 — two days in, work is being filed) and the deterministic engine's code-evidence join checked 2 board-done issues, finding 0/2 code-corroborated. Worklist: 5 deterministic + 3 editorial findings.

Prior findings re-tested:

  • F-1 coverage-gap (Google Calendar / BOT-4482) — STILL OPEN. Baghiz Firdausa (named owner of the Sprint Goal 1 work) still shows 0 issues in sprint 21473.
  • F-2 coverage-gap (Impact Report / BOT-4568) — STILL OPEN, changed: 1 issue now exists under the Epic, but it resolves to a non-dev discipline — still 0 BE/FE implementation tasks on the board.
  • F-3 (BOT-4574 Epic mis-link) — RESOLVED. Not re-surfaced by this pass's engine run (which computes Epic coverage directly); no BOT-4602 undercount or Linggar-idle misreport appeared this time.
  • F-4 (Eko Aprianto zero Jira issues) — RESOLVED. The engine now finds Eko Aprianto assigned in Jira (an assignment-drift row surfaced instead — see F-12 below: the repo's staffing mirrors hadn't caught up).
  • F-5 (BOT-4621 unassigned/unsprinted) — RESOLVED. Not re-surfaced this pass.
  • F-6 roster-mismatch (Tito/Izzul initiative-QA rows unfiled) — STILL OPEN. Izzul still shows 0 issues in sprint 21473.
  • F-8 (upstream engine norm() dotted-slug bug) — STILL OPEN (upstream, unverified — no way to confirm the fix landed from this repo).
  • F-9 (AI Agent Testing task breakdown "exists in Jira only") — RESOLVED, corrected: the premise itself was stale. chatbot/ai-agent-testing/rfcs/historical-validation.task-breakdown.md already exists in the repo (Phase 1/2 chunks, effort-estimated per task). The sprint doc's Initiative-index row and ## Deviations note ("no task breakdown yet") are what's stale, not the repo — see F-16 below. The live board now shows 11 sprint-scoped issues against BOT-3351 (1 done, 2 in-progress), consistent with the breakdown's task numbering (BOT-4618 joins to Task 2.1).
  • F-10 (BOT-4084 off-plan item) — RESOLVED. Not re-surfaced this pass.

New findings from the 2026-07-15 pull:

  • F-12 assignment-drift (deterministic): chatbot/engineering-q3-2026/README.md implementors: lists only Agus, but Jira shows Eko Aprianto and Samudra Linggar also assigned under Epic BOT-4620. Applied this pass: added both to README implementors:, the roadmap Implementors column, and capacity.md Allocation rows (as engineering-q3-2026 (BE/FE) TBD, matching the existing TBD convention for Agus/Tito on this bucket). lint:delivery re-run clean after the edit.
  • F-13 status-progress (editorial, informational): AI Agent Testing shows 1 done / 2 in-progress of 11 sprint issues, but the 1 done issue (BOT-4618) has 0/1 code-corroborated (see F-14) — so progress_done_days/progress_pct correctly stay at 0% in chatbot/ai-agent-testing/delivery/status/2026-07-14.md. No status/ edit needed; documenting only.
  • F-14 code-evidence-gap: BOT-4618 ([BE] Task 2.1 — test case name column, processing-stat) is marked Won't Fix on the board with no merged-code trace found in chatbot/chatbot-fe (search partial — qontak-designer/hub not checked). A "Won't Fix" status is plausibly code-free by definition, but this is a board-status call, not a repo edit — routed to the assignee/DRI to confirm.
  • F-15 code-evidence-gap: BOT-4653 ([BE] reduce payload to Ably in bot-preview) is marked Done with no merged-code trace in chatbot/chatbot-fe (partial — notification-service/qontak-designer not checked). Unlike F-14 this is claimed Done, not Won't Fix — 0 man-days corroborated until a merge/PR reference is found. Routed to the assignee/DRI.
  • F-16 (new, stale-doc): the sprint doc's own P3 row ("no task breakdown yet") and ## Deviations note are now factually wrong — the breakdown exists both in the repo and (per F-9) is already reflected in 11 live Jira issues. Owner: TPM — correct the Initiative-index "Ready basis / effort context" cell and Deviations prose on next plan touch (routed to sprint-planner amend mode, not a resync rewrite of the plan body).

This pass DID apply the F-12 staffing-mirror fix (README + roadmap + capacity, lint:delivery clean). It did not file any Jira ticket, reassign/re-link any issue, or touch the sprint doc's plan body beyond the findings ledger (F-16 is routed to sprint-planner, not applied here).

Downstream propagation worklist​

#FindingCanonical artifact to reconcileOwning tool / skill
1F-1 Google Calendar zero board presence (still open)File BOT-4482-scoped Tasks for Agus/Baghiz/Tito into sprint 21473sprint-jira-sync (sprint mode)
2F-2 Impact Report 0 BE/FE dev work (still open)File BOT-4568-scoped BE/FE Tasks into sprint 21473sprint-jira-sync (sprint mode)
3F-6 Izzul QA rows unfiled (still open)File QA drafting Tasks under BOT-4602/BOT-4568/BOT-3351 for Izzul/Titosprint-jira-sync
4F-8 engine false-positive (still open, upstream)No repo edit; report norm() gap to qontak-docsUpstream issue against sprint-reconcile.mjs
5F-14 BOT-4618 Won't Fix, no code traceConfirm code-free status or find the merge; DRI callassignee / DRI board fix
6F-15 BOT-4653 Done, no code traceFind the merge/PR reference or revert board statusassignee / DRI board fix
7F-16 stale P3 "no task breakdown yet" doc textAmend Initiative-index cell + Deviations prosesprint-planner (amend mode)

Always run npm run lint:delivery (DOCS_ROOT pointed at this repo) after any propagation edit.