Skip to content

Jira issue-tracking integration

Area: process / project tooling / tracking Raised during: PM session 2026-06-25 (documentation hosting & access-control discussion) Jira: (this item bootstraps the project — add the key once the project exists) Priority: Medium now, rising to High as the implementation phase deepens Start trigger: Kick off when Sprint 1 (Images pilot) closes — i.e. at the Sprint 1 acceptance gate. Not before; Sprint 1 keeps using the git deferred-note workflow. (PM decision 2026-06-25.) Owner: Yuji Shinozaki

Context

Confluence and Jira Cloud are now available to the Mandala team (they were not when the git/mkdocs interim workflow was set up). Of the two halves of the documentation/tracking strategy, Jira is the more urgent: as the project moves from spikes/planning into sustained implementation, ad-hoc deferred notes and session logs stop being sufficient for tracking who-owns-what and what's in-flight. The Confluence/docs-hosting half is comparatively long-run and is parked as a low-priority spike — see Spike 9.

The deferred-note format already anticipated this: every note carries a **Jira:** (add when available) placeholder for backfill, and docs/deferred/README.md states each file should map 1:1 to a ticket once Jira exists.

What needs deciding / doing

  1. Stand up the Jira project — key, issue types, workflow. Decide how it relates (if at all) to any existing UVA Library Jira projects.
  2. Deferred notes → tickets (1:1). Backfill the current ~14 open deferred notes as tickets and record each ticket key back in the note's **Jira:** header. Decide whether the markdown note or the Jira ticket is the source of truth going forward (recommend: ticket for status/assignment, markdown for the detailed technical context, cross-linked).
  3. Spikes ↔ Jira. Decide whether spikes get tracked as Jira epics/tasks or stay git-only with the spikes-plan as the index.
  4. Define the working rhythm — what gets a ticket vs. what stays a session log; how session-end deferred notes flow into Jira.
  5. Session logs stay in git (not Jira) — they are narrative, not trackable work items.

When to start

Sequenced after Sprint 1 (Images pilot) completes — start at the Sprint 1 acceptance gate so the first sprint isn't disrupted mid-flight. The Confluence/docs-hosting half (Spike 9) trails this further still; it can wait longer. When Sprint 1 closes, promote this note to High and begin step 1 below.

Why this is separable from the docs-hosting spike

Jira adoption delivers value immediately and independently: it does not depend on solving the markdown → Confluence sync, the public/private repo split, or the git submodule question. It can proceed now while the hosting spike waits.