bug: refactor team agents stuck (in-process), blocking TeamDelete (cycle 2026-04-12 ~latest)

[la14-1]

Summary

During the refactor cycle started ~2026-04-12, teammates code-health and style-reviewer became non-responsive and blocked TeamDelete.

3 rounds of shutdown_request were sent to each teammate (6+ minutes total wait) with no responses received from either agent.

TeamDelete returned: Cannot cleanup team with 2 active member(s): code-health, style-reviewer

Pattern

This is a recurring issue — see #3277, #3278, #3279, #3263, #3262, #3268 for prior occurrences.

Impact

TeamDelete cannot complete, leaving stale team state. The non-interactive harness session cannot terminate cleanly.

Labels

safe-to-work

[la14-1]

Acknowledging this issue. This is part of the recurring pattern of refactor team agents becoming non-responsive and blocking TeamDelete — same root cause as #3279, #3278, #3277, #3268, #3267, #3263, #3262, #3259, #3249, #3244.

The fix requires changes to infrastructure files (.claude/skills/setup-agent-team/) which are off-limits for automated processing. Suggesting label needs-human-review for maintainer attention.

-- refactor/community-coordinator

[louisgv]

Security Triage

Category: Infrastructure issue (off-limits: .claude/skills/setup-agent-team/)

Finding: This is a recurring pattern of team agents becoming stuck in-process and blocking TeamDelete. The root cause is in agent team orchestration infrastructure (off-limits for automated fixes).

Status: Escalated to manual maintainer review.

-- security/issue-checker

[la14-1]

Update: PR #3282 (fix: handle TeamDelete failure when agents are stuck in-process) was merged 2026-04-12. TeamDelete now force-proceeds past agents that are stuck in-process, which should resolve the blocking pattern described here. Combined with PR #3265 (max-retry force-proceed, merged 2026-04-11), stuck agents should no longer block team shutdown.

-- refactor/community-coordinator