Bot temporarily unavailable #3935

Activity

1. worktrunk-bot

   added

   tend-outageTracks bot outage incidentsTracks bot outage incidents

   on Aug 27, 2026
2. worktrunk-bot commented on Aug 27, 2026

   

   worktrunk-bot

   on Aug 27, 2026

   CollaboratorAuthor

   More actions

   When	Run	Trigger
   2026-08-27T16:26:46Z	workflow run	#3936
3. worktrunk-bot commented on Aug 27, 2026

   

   worktrunk-bot

   on Aug 27, 2026

   CollaboratorAuthor

   More actions

   When	Run	Trigger
   2026-08-27T16:57:16Z	workflow run	N/A
4. worktrunk-bot commented on Aug 27, 2026

   

   worktrunk-bot

   on Aug 27, 2026

   CollaboratorAuthor

   More actions

   When	Run	Trigger
   2026-08-27T20:27:05Z	workflow run	N/A
5. worktrunk-bot commented on Aug 27, 2026

   

   worktrunk-bot

   on Aug 27, 2026

   CollaboratorAuthor

   More actions

   When	Run	Trigger
   2026-08-27T20:39:48Z	workflow run	#3937
6. worktrunk-bot commented on Aug 27, 2026

   

   worktrunk-bot

   on Aug 27, 2026

   CollaboratorAuthor

   More actions

   When	Run	Trigger
   2026-08-27T21:47:50Z	workflow run	#3913
7. worktrunk-bot commented on Aug 27, 2026

   

   worktrunk-bot

   on Aug 27, 2026

   CollaboratorAuthor

   More actions

   When	Run	Trigger
   2026-08-27T22:55:08Z	workflow run	#3938
8. worktrunk-bot commented on Aug 27, 2026

   

   worktrunk-bot

   on Aug 27, 2026

   CollaboratorAuthor

   More actions

   When	Run	Trigger
   2026-08-27T23:38:23Z	workflow run	#3939
9. worktrunk-bot commented on Aug 27, 2026

   

   worktrunk-bot

   on Aug 27, 2026

   CollaboratorAuthor

   More actions

   When	Run	Trigger
   2026-08-27T23:42:32Z	workflow run	#3940
10. worktrunk-bot commented on Aug 27, 2026

    

    worktrunk-bot

    on Aug 27, 2026

    CollaboratorAuthor

    More actions

    When	Run	Trigger
    2026-08-27T23:45:19Z	workflow run	N/A
11. worktrunk-bot commented on Aug 28, 2026

    

    worktrunk-bot

    on Aug 28, 2026

    CollaboratorAuthor

    More actions

    Drained by the tend-review-runs sweep. Cause found and already resolved; two rows need a fresh trigger, noted below.

    Cause

    All ten rows are the same failure: tend 0.1.19 discarded the runner's PATH across the sandbox privilege boundary, so .config/tend.yaml's sandbox_setup: line uv tool install --quiet pre-commit died with bash: line 4: uv: command not found (exit 127) and the agent never booted.

    Clean bisection on the action pin — 0.1.18 green at 15:14Z, all ten failures on 0.1.19, green again on 0.1.20 at 23:51Z:

    Run max-sixty/tend/claude@0.1.18   # 33086703145, 15:14Z, success
    Run max-sixty/tend/claude@0.1.19   # 33127253361, 23:45Z, failure (last of ten)
    Run max-sixty/tend/claude@0.1.20   # 33127767243, 23:51Z, success
    

    This is the upstream regression fixed by max-sixty/tend#1071 ("Preserve setup-added tool paths in the Claude sandbox") and shipped in 0.1.20, which landed here in #3940 at 23:51Z. No adopter-side change is needed — .config/tend.yaml and .github/actions/tend-setup/action.yaml were both correct throughout, and main has been on 0.1.20 since.

    Row status

    Run	Workflow	Trigger	Status
    33087715330	tend-triage	#3934	Needs a fresh trigger
    33093061373	tend-review	#3936	Moot — PR merged
    33095646749	tend-notifications	cron	Recovered on the 23:51Z tick
    33113154647	tend-notifications	cron	Recovered on the 23:51Z tick
    33114235671	tend-triage	#3937	Recovered — a later run answered it at 00:01Z
    33119553121	tend-mention	#3913	Recovered — a later run answered it at 00:01Z
    33124133043	tend-review	#3938	Moot — PR merged
    33126810796	tend-review	#3939	Needs a fresh trigger
    33127036034	tend-review	#3940	Moot — PR merged
    33127253361	tend-notifications	cron	Recovered on the 23:51Z tick

    Why the two remaining rows were not re-run

    gh run rerun would reproduce the failure rather than drain it. A re-run uses the workflow file from the original run's commit, and that commit pins the broken version:

    $ gh api repos/max-sixty/worktrunk/actions/runs/33087715330 --jq '{event, head_sha}'
    {"event":"issues","head_sha":"37141b8eb10bfafde1d127f497545b8c621a7409"}
    
    $ git show 37141b8eb:.github/workflows/tend-triage.yaml | sed -n '49p'
          - uses: max-sixty/tend/claude@0.1.19
    

    So each re-run would fail the same way and file a fresh row here, on an incident that is already over. What recovers each one instead:

    • Wrong template values when is using wt switch pr:<number> #3934 (robsonpeixoto's report, opened 15:25Z, still at zero comments) — tend-triage fires only on issues: [opened], so nothing can re-fire it. An @worktrunk-bot mention on the issue is the only path.
    • Color homepage command comparison #3939 — tend-review includes reopened in its trigger types, so a push or a close/reopen re-fires it. It's your PR, so that's your call rather than something this sweep should do.

    Closing so the next outage opens a clean incident rather than folding into this one. The gh run rerun gap is reported upstream on max-sixty/tend#1066.
12. worktrunk-bot

    closed this as completedon Aug 28, 2026