Runbook: update the public Jira area
Purpose
Update the public Jira area in this repo:
- DE:
/de/jira/,/de/jira/sprint/,/de/jira/backlog/,/de/jira/historie/,/de/jira/erledigt/,/de/jira/issues/J01-*/ - EN:
/en/jira/,/en/jira/sprint/,/en/jira/backlog/,/en/jira/history/,/en/jira/erledigt/,/en/jira/issues/J01-*/ - Step pages only for subtasks with public details:
/de/jira/issues/<PARENT>/steps/<SUBTASK_KEY>/and/en/jira/issues/<PARENT>/steps/<SUBTASK_KEY>/ - Filtered Jira remote links are shown as relative links in the public Jira area.
Requirements
- Access to the
atlassianCLI (tool owner configured locally) andjq - Write access in the repo
- Public policy: no Jira Cloud links, no email addresses
Path Semantics
work/...is the public source path for Jira work docs in this repo.- Public URI paths live under
/de/...and/en/.... - A work doc can therefore live at the filesystem path
work/jira/J01-106/index.mdwhile being published at the URI/de/jira/issues/J01-106/. - The mapping is created by
permalinkand the generator/Jekyll flow, not by a 1:1 folder mirror.
Steps
1) Update issues and statuses in Jira (SSOT for issue state).
- For sprint work, also verify the active sprint, set sprint labels, and update the public sprint docs.
- Keep only one canonical Jira remote link per public work doc; DE/EN language split stays a GitHub Pages concern.
2) Normal mode for known changes:
DOCS_REPO=$PWD sh ../.agents/skills/lebenslauf-web-vorlage/skills/jira-state-sync/scripts/run-sync.sh jira-change --change J01-95:summary
DOCS_REPO=$PWD sh ../.agents/skills/lebenslauf-web-vorlage/skills/jira-state-sync/scripts/run-sync.sh pages-change HEAD~1
DOCS_REPO=$PWD sh ../.agents/skills/lebenslauf-web-vorlage/skills/jira-state-sync/scripts/run-sync.sh resume
If a targeted run would regress untouched keys before writing, the guard stops the run. Treat that as a gap in the journaled sync; a deliberate full refresh is required afterwards.
3) Repair mode / full refresh:
DOCS_REPO=$PWD sh ../.agents/skills/lebenslauf-web-vorlage/shared-tooling/jira-pages/update-public-jira-pages.sh
Optional (full refresh):
DOCS_REPO=$PWD sh ../.agents/skills/lebenslauf-web-vorlage/shared-tooling/jira-pages/update-public-jira-pages.sh --full
4) Verify GitHub Pages targets from Jira locally:
DOCS_REPO=$PWD sh ../.agents/skills/lebenslauf-web-vorlage/shared-tooling/jira-pages/verify-public-jira-pages.sh
Optional: audit legacy Confluence targets directly:
DOCS_REPO=$PWD sh ../.agents/skills/lebenslauf-web-vorlage/shared-tooling/jira-pages/verify-public-jira-pages.sh --legacy-confluence-audit
The URL conversion from Jira follows these rules:
https://docs.template.ysdani.com/...becomes a site-relative URI- neutral pages remain only
/or/index.html - language-bound pages are verified as
/de/...in DE context and/en/...in EN context - verification runs against this local GitHub Pages repo
- the legacy audit reports old
atlassian.net/wikitargets asLEGACY_CONFLUENCE<TAB>KEY<TAB>LINK_ID<TAB>URL<TAB>TITLE
5) Spot check:
- The entry page under
jira/points to backlog, history, and sprint board. - Backlog lists show each listed issue’s status directly in the same line.
- The history view bundles sprint history and globally visible completed top-level issues.
erledigt/only remains as a compatibility path to the history view.- Issue and step pages under
jira/expose short, visible cross-paths to parent pages and public steps; issue families with multiple public pages additionally show a contents key. - Subtasks without public details do not get their own page.
- The base record of a subtask is step number, title, and status.
- Pure step-page metadata such as parent, step number, status, key, or updated timestamp do not count as additional details.
- Subtasks with additional public details do get a step page.
- Additional public details are the subtask’s own domain content or canonical links beyond the base record and metadata.
- This includes, in particular, a public work doc under
the filesystem path
work/jira/J01-<KEY>/, a canonical Jira remote link to its public URI, the subtask’s own evidence, its own closure statement, or other canonical target pages. - If a public step page under
jira/issues/<PARENT>/steps/<SUBTASK_KEY>/is used instead of a public issue page, the same expectation applies to canonical linking and carry-over. - When new step pages are introduced for existing subtasks, existing public mentions are updated where they directly touch that step page or its parent issue.
- If a public issue or step page exists, verification checks it against Jira remote links, legacy paths, and duplicate detail URLs.
6) Hygiene:
- No
atlassian.netlinks in the output - No email addresses in the output
- Remote links are rendered as relative site links
- Remote links follow the language context of the page
(
/de/...on DE pages,/en/...on EN pages) - The sprint board groups current sprint work by category labels
(
sprint-goal,sprint-support,sprint-admin,sprint-unplanned) - The sprint board only shows the same top-level issues as the Jira board; steps stay visible on parent and step pages.
- Backlog and history show the status on every listed issue line.
- Only generated EN Jira pages under
en/work/jira/are synchronized after the DE render
Rollback
- Revert the affected files and generate the public Jira area again.
Monitoring
- Sprint board, backlog, and history match Jira status categories and target paths.
- The sprint board has dedicated row groups for sprint categories.
- Backlog lists show each listed issue’s status directly in the same line.
erledigt/only points to the history view and no longer carries its own full listing.- Issue and step pages carry an
Updatedtimestamp from Jira. gitremains the reliable history: diffs only appear when content changes.shared-tooling/jira-pages/verify-public-jira-pages.shverifies imported GitHub Pages targets locally for both DE and EN.shared-tooling/jira-pages/verify-public-jira-pages.shalso verifies the absence of public legacy paths and duplicate detail URLs.shared-tooling/jira-pages/verify-public-jira-pages.sh --legacy-confluence-auditreports remaining legacy Confluence remote links with Jira key and link ID..local/jira-sync-cache/and.local/jira-sync-journal/keep the local snapshot and resume state for the journaled normal mode.- Guard failures from targeted runs are also recorded in
.local/jira-sync-journal/; those runs do not write public Jira files and require a full refresh afterwards. - This docs repo does not use a generic
dev-/previewGit flow; after local checks, changes are integrated directly intomain.