[guide] LAUNCH.md readiness gate cites closed tracking issues; refresh-community-people still red and its live blocker (#568) is not referenced

[hivecommons-hive]

Documentation Gap

The launch-readiness gate table in LAUNCH.md ("Launch-readiness gate: zero-red content pipelines") lists tracking issues that no longer reflect reality:

• Refresh community profiles (refresh-community-people.yml) is tracked as "[ci-maintainer] Refresh community profiles fails weekly: repo blocks GITHUB_TOKEN from creating PRs #122", but [ci-maintainer] Refresh community profiles fails weekly: repo blocks GITHUB_TOKEN from creating PRs #122 was closed 2026-09-23 as NOT_PLANNED (duplicate of [ci-maintainer] Import reference architectures fails at create-pull-request: GITHUB_TOKEN blocked from creating PRs (same root cause as #122) #519, itself closed COMPLETED 2026-09-21). Meanwhile the pipeline's most recent scheduled run (2026-09-21, run 35555457167) failed — at the build step with HTML minifier errors, a different failure mode than either closed issue describes. The current open tracker for this workflow's automation blocker is [scanner] Automation PRs blocked by DCO: create-pull-request steps never set signoff: true (import-architectures, refresh-community-people) #568 (DCO signoff: true missing on the create-pull-request step), which is referenced nowhere in LAUNCH.md, MILESTONES.md, or ROADMAP.md (verified by grep).
• Import reference architectures is tracked as "[ci-maintainer] Import reference architectures fails daily: raster-embedded SVGs, rsvg-convert missing (fix PR #49 open) #121" (closed COMPLETED); the daily run went green on 2026-09-24, so that row happens to be accurate today, but the table gives no signal about where to look when it regresses.

A maintainer checking the W-1 gate ("2 consecutive scheduled runs green") who follows the "Tracking: #122" link lands on a closed-as-duplicate issue and then a closed-COMPLETED issue, and could reasonably conclude the weekly pipeline is handled — it is not. The gate table currently hides the red pipeline instead of surfacing it.

Impact

Launch coordinators and maintainers. The readiness gate is the go/no-go check for the KubeCon NA 2026 launch (2026-11-09); a tracking column that points at closed issues for a still-failing pipeline defeats its purpose. Per LAUNCH.md's own "How to use this file", the table is supposed to be updated when execution drifts.

Recommendation

Update the readiness-gate table's Tracking column:

• Refresh community profiles: point at the open blocker [scanner] Automation PRs blocked by DCO: create-pull-request steps never set signoff: true (import-architectures, refresh-community-people) #568 (and note the latest failure mode — build-step HTML minifier errors — if it persists after the DCO fix).
• Consider adding a "last green run" or "status as of" column, or a line under the table stating the current consecutive-green count per pipeline, so the gate is checkable without cross-referencing closed issues.

Evidence: gh issue view 122 (closed 2026-09-23, NOT_PLANNED, dup of #519); gh issue view 519 (closed COMPLETED); gh run list --workflow refresh-community-people.yml (last 3 scheduled runs 09-07/09-14/09-21 all failed); gh run view 35555457167 --log-failed (HTML minifier errors in build step); gh issue view 568 (open); grep of LAUNCH.md/MILESTONES.md/ROADMAP.md for "568"/"519" returns nothing.

---

Filed by guide agent (ACMM L4 — issues-only mode)

🐝 Hive Agent: guide | Instance: hosted-available-lke648397-260827-5n31 | SHA: unknown

— hive: agent=guide backend=copilot model=kimi-k3 copilot=1.0.88