[ci-maintainer] v5 Top-up conflicting again: 9+ failures since 20:43Z — v4 ad89567e vs v5 84a86b30, 17 files, recurred after #7450 layout adoption

[kubestellar-hive]

CI Issue

The v5 Top-up workflow is failing again on every v4 push with "v4 -> v5 forward-merge conflicts and needs a human". This is a recurrence after #7337 was closed (16:59Z) and after the structural fix #7450 (layout/v5-adopt-v4, tracker #7369) merged into v5.

No sync/v4-to-v5 branch exists and no open PR covers the resolution.

Evidence

Failing runs (workflow: v5 Top-up, all on v4 workflow_run), 9 failures 20:43Z–22:06Z on 2026-09-17:
35280346013, 35279734874, 35279463895, 35278365817, 35275512843, 35274965029, 35273896575, 35273225626, 35272497427

Reproduced locally at v4 ad89567e (PR #7461, release-gate/v4.52.0) vs v5 84a86b30 (PR #7450): git merge --no-ff origin/v4 conflicts in 17 files:

src/cmd/hive/main.go
src/docs/agent-configuration.md
src/docs/api-reference.md
src/docs/design/master-delivery-wrapped.md
src/pkg/agent/manager.go
src/pkg/agent/manager_kick.go
src/pkg/agent/manager_tmux.go
src/pkg/agent/tmux_coverage_test.go
src/pkg/config/config.go
src/pkg/dashboard/api_contribute.go
src/pkg/dashboard/api_governor.go
src/pkg/dashboard/static/index.html
src/pkg/dashboard/status_builder.go
src/pkg/github/client.go
src/pkg/github/stale_draft_test.go
src/pkg/hub/saas.go
src/pkg/scheduler/scheduler.go

Notably main.go, config.go, manager.go, saas.go still conflict even though #7450 adopted the v4 layout on v5 — v4 has advanced past the adoption point.

Recommendation

A human (or merge-authorized agent) needs to resolve per the workflow's own instructions:

git checkout -B sync/v4-to-v5 origin/v5
git merge --no-ff --signoff origin/v4
# resolve, then push and open a PR against v5

Conflict resolution touches production source, which is outside the ci-maintainer remit, so no PR accompanies this issue. Until resolved, every v4 push produces a red v5 Top-up run.

---

Filed by ci-maintainer agent (ACMM L4/L5 — hold-gated mode)

🐝 Hive Agent: ci-maintainer | Instance: hosted-available-oke-11-placeholder-r05x | SHA: unknown

— hive: agent=ci-maintainer backend=copilot model=claude-fable-5

[clubanderson]

Note for whoever picks this up: the top-up is already open as #7494 (topup/v5-from-v4-c -> v5), so I declined to open a duplicate and reviewed it instead (#7494 (comment)).

Short version of that review: #7494's conflict resolutions look semantically sound — I re-resolved the same merge independently and cross-checked a sentinel symbol for every v5 divergence touching a conflicted file, and none of v5's commits were silently reverted (the legacy tmux layer and the dead trust-tier helpers correctly stayed deleted rather than being resurrected from v4's side). But it tops up only through 82308e22f, and v4 has since advanced by bf1c4b6b (#7451/#7460) and c80eaa69 (#7453/#7457) — SetHubPushedDashboardURL and handleAuthReturn are absent from that branch. Both land in files that conflicted, so the Top-up workflow will fail again on the same paths after #7494 squash-merges. This issue probably shouldn't be closed until a top-up reaching current v4 HEAD lands.

[clubanderson]

Picking this up. #7494 landed v5 through 82308e22f; #7513 (on the workflow's own sync/v4-to-v5 branch) carries v5 to current v4 HEAD b2c98f6bb (15 commits) with a proper two-parent merge commit. Once it merges I'll dispatch v5 Top-up to confirm it reports level/green, then close this.

[clubanderson]

Closing. The conflict this issue reports is structural, not a one-off: v4 agents are continuously extracting domains out of god files (#7278, #7303, #7435, #7491, #7238), so a textual git merge origin/v4 into v5 conflicts on nearly every round and v5-topup.yml correctly fails loudly.

What was done instead of waiting for a clean auto-merge:

• v5 adopted v4's file layout (🌱 v5: adopt v4's file layout for cmd/hive, pkg/config, pkg/agent, pkg/hub #7450) so the drift is now declaration-level, not file-level.
• Six hand-resolved top-ups on the workflow's own sync/v4-to-v5 branch this session: 🌱 v5: top-up from v4 through 13ef3e6e1 (44 commits) #7466, 🌱 v5: top-up from v4 through cfc3a6989 (11 commits, v4.52.2) #7481, 🌱 v5: top-up from v4 through 82308e22f (12 commits, v4.53.0) #7494, 🌱 Forward-merge v4 into v5 through b2c98f6bb (15 commits) #7513, 🌱 Forward-merge v4 into v5 through f53a076ff (5 commits) #7525, 🌱 Forward-merge v4 into v5 through ff7630e1f (6 commits) #7541, and 🌱 Forward-merge v4 into v5 through 201c69c8a (20 commits) #7545 (in CI) — v5 is level with v4 @ 201c69c (v4.55.1) once 🌱 Forward-merge v4 into v5 through 201c69c8a (20 commits) #7545 lands.
• Procedure: declaration-level 3-way merge per conflicting package, v5-only seams preserved, full go test ./... before push.

v5-topup.yml stays as the alarm: when it fails again, the ci-maintainer agent re-files and the same procedure applies. Keeping this specific instance open adds nothing beyond that.

[github-actions]

@kubestellar-hive[bot] this issue was closed by @clubanderson as completed.

If this was closed by a merged PR, the fix only takes effect once it ships in a tagged release and your deployment updates to it — check your running version against the fix before concluding the fix failed.

If you are still seeing the problem, comment /reopen and it will be reopened for you — you do not need write access. Please add what you are still observing (running version, logs, steps) so it is not closed the same way again.