Default branch sweep - a repository we own is not green

[github-actions]

3 of 26 PUBLIC repositories we own do not have a genuinely green default branch.

This run saw no private repository. A repository-scoped GITHUB_TOKEN does not list them at all, so they are not skipped-and-reported here -- this job never learns they exist. Set an OA_SWEEP_TOKEN secret holding a token with actions:read on the private repositories to cover them.

cancelled, timed_out and startup_failure are not green: such a run proves nothing about the tree. The usual cause is concurrency cancel-in-progress when several merges land at once, so re-run the cancelled run rather than assuming the older success still stands.

This issue is rewritten in place once a day and closed when everything is green. Editing a body does not notify, so a long-lived gap does not become a daily ping.

OpenAdaptAI/openadapt-desktop (main)

• Notify docs — newest run concluded failure on 18a3198; the workflow file changed afterwards (2026-08-02T17:45:03Z), so re-run it to find out whether the fix took (run)
• pip in /. - Update #1498972175 — newest run concluded failure on ac01e90 (run)
• Build managed FFmpeg runtime: not-push-gated — dispatch or schedule only, so the failure run says nothing about the tree (retired 2026-07-26T04:10:38Z, after that run)

OpenAdaptAI/openadapt-evals (main)

• evidence-freshness — newest run concluded failure on 9dd8021 (run)

OpenAdaptAI/openadapt-ops (main)

• Production DB logical backup — newest run concluded failure on 9da0e4d AFTER the workflow was retired (2026-08-02T16:30:57Z), so it failed as currently configured (run)
• CodeQL: stale-green — success, but it tested 9da0e4d and head is 0188629 -- that pass is evidence about a tree nobody runs any more
• Default branch sweep: in-flight — in_progress for 0.0h on 0188629 -- not yet evidence either way
• Docs CI: stale-green — success, but it tested 74e8146 and head is 0188629 -- that pass is evidence about a tree nobody runs any more
• Secret Scan: stale-green — success, but it tested f2b1daf and head is 0188629 -- that pass is evidence about a tree nobody runs any more
• Sync Documentation: stale-green — success, but it tested 9da0e4d and head is 0188629 -- that pass is evidence about a tree nobody runs any more

Green, with notes — none of these is a failure

retired and not-push-gated entries are frozen history: GitHub keeps a workflow's last run for ever, so a workflow that stopped running would otherwise be reported red permanently. stale-green passed on an older tree. in-flight has been running for less than 6h.

• OpenAdaptAI/OpenAdapt
  • Noncanonical docs guard: stale-green — success, but it tested 3771129 and head is f8369b2 -- that pass is evidence about a tree nobody runs any more
  • Notify docs: stale-green — success, but it tested 76595ac and head is f8369b2 -- that pass is evidence about a tree nobody runs any more
• OpenAdaptAI/openadapt-flow
  • Paper: stale-green — success, but it tested ccdd155 and head is ddbbd64 -- that pass is evidence about a tree nobody runs any more
• OpenAdaptAI/openadapt-grounding
  • Release and PyPI Publish: not-push-gated — dispatch or schedule only, so the failure run says nothing about the tree (retired 2026-07-28T02:22:09Z, after that run)
• OpenAdaptAI/openadapt-herald
  • weekly-digest: retired — workflow disabled_inactivity; the failure run is frozen history
• OpenAdaptAI/openadapt-retrieval
  • Release and PyPI Publish: not-push-gated — dispatch or schedule only, so the failure run says nothing about the tree (retired 2026-07-28T02:22:12Z, after that run)
• OpenAdaptAI/openadapt-viewer
  • Release and PyPI Publish: not-push-gated — dispatch or schedule only, so the failure run says nothing about the tree (retired 2026-07-28T02:22:16Z, after that run)

Swept 26 repositories in 122 API reads. https://github.com/OpenAdaptAI/openadapt-ops/actions/runs/30897812208