Dashboard deploy awaiting manual approval since 2026-09-23T22:49Z (2 runs queued)

[fro-bot]

Summary

Multiple Deploy Dashboard (deploy-dashboard.yaml) runs are stuck at the dashboard environment's required-reviewer approval gate and have not been approved or rejected. dashboard.fro.bot is now five versions behind the newest dispatched build.

Details

Run	Target version	Dispatched	Status
35930438231	2026.09.20	2026-09-23T22:49:36Z	still waiting
36008420862	2026.09.21	2026-09-24T13:50:10Z	cancelled (queue auto-cancel)
36298026416	2026.09.22	2026-09-27T05:43:33Z	cancelled (queue auto-cancel)
36325597185	2026.09.23	2026-09-27T14:21:09Z	cancelled (queue auto-cancel)
36455658035	2026.09.24	2026-09-28T17:05:24Z	pending, queued behind 35930438231, unactioned ~10.6 hours

The last successful dashboard deploy remains 35646229785 (version 2026.09.19, completed 2026-09-21T19:39:53Z). apps/dashboard/docker-compose.yaml on main is still pinned to 2026.09.19 (the pin commit is written only after a successful deploy), confirming the running production image has not advanced past that version even though five newer versions (2026.09.20–2026.09.24) have since been dispatched.

This is the same "stranded behind an approval gate" pattern flagged for CLIProxy (#1437), VPN (#1436), and Gateway (#1412) — a deploy run sits in the waiting/pending state (not failed) until a human approves or rejects it in the dashboard GitHub Environment, and unapproved runs are auto-cancelled only when a newer dispatch queues behind the original waiting run.

Current health

• https://dashboard.fro.bot/api/healthz returns 200 — the app is up and serving the last successfully deployed version (2026.09.19).
• No action was taken automatically; approval gates are intentional and this workflow does not approve or redeploy.

Suggested action

Visit the Actions tab and approve or reject the oldest queued dashboard environment deployment (35930438231, 2026.09.20) and the newest (36455658035, 2026.09.24). Approving or rejecting the oldest first is required before newer queued runs can execute.

---

Filed by the daily autoheal run (category 7: Deploy Pipeline Health). Reporting only — no code or deploy changes were made. Updated 2026-09-29 to reflect current queue state.

[fro-bot]

Update — 2026-09-28 (daily autoheal)

Still stranded, and the queue has grown by two more entries since this issue was filed:

• Oldest unactioned: 35930438231 (target 2026.09.20) — deploy-dashboard still waiting, unactioned for ~4 days 5 hours.
• 36008420862 (target 2026.09.21) — cancelled (superseded in queue).
• 36298026416 (target 2026.09.22) — cancelled (superseded in queue).
• Newest queued: 36325597185 (target 2026.09.23) — pending, unactioned for ~13.5 hours.

No later successful deploy-dashboard run exists; production remains on 2026.09.19 (last success: 35646229785, 2026-09-21T19:39:53Z). apps/dashboard/docker-compose.yaml pin on main is still 2026.09.19, consistent with no successful deploy since.

https://dashboard.fro.bot/api/healthz returns 200 — the app is up and serving 2026.09.19.

Reporting only — no redeploy or approval action taken.

Filed by the daily autoheal run (category 7: Deploy Pipeline Health).