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

[la14-1]

Description

community-coordinator and pr-maintainer agents spawned during the 2026-04-12 refactor cycle became stuck in-process and never responded to shutdown_request messages (3 rounds sent), blocking TeamDelete.

Repro

1. Spawn refactor team with community-coordinator and pr-maintainer teammates
2. Agents go in-process (never idle, never send messages back to team-lead)
3. Send shutdown_request 3 times to each — no response
4. TeamDelete fails: 'Cannot cleanup team with 2 active member(s)'

Impact

TeamDelete could not complete. Worktrees cleaned up manually via git worktree prune.

Related

Same pattern as #3244, #3249, #3256, #3259, #3262, #3263, #3267, #3268

[la14-1]

Acknowledged. This is the same class of stuck-agent issue as #3267, #3263, #3262, #3244 — all caused by agents blocking in-process during TeamDelete. The root fix (prompt-level) lives in off-limits files (.claude/skills/setup-agent-team/). Nothing actionable for the automated refactor team this cycle.

-- 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. This directly addresses the TeamDelete blocking described here by force-proceeding past stuck agents.

-- refactor/community-coordinator

[louisgv]

Updated Triage Assessment

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

Severity: HIGH

Update: The developer (la14-1) reports that this has been addressed by merged fixes:

• PR fix: handle TeamDelete failure when agents are stuck in-process #3282: TeamDelete now force-proceeds past agents stuck in-process
• PR fix: add max-retry force-proceed to prevent infinite shutdown loop #3265: Max-retry force-proceed added to prevent infinite shutdown loops
• PR fix: replace plan_mode_required with message-based approval in refactor team #3257: Message-based approval system

Recommended action: Monitor next refactor cycles to verify the fix resolves the stuck-agent pattern. If agents still become stuck after these fixes, the root cause requires deeper investigation in the team orchestration layer.

Status: safe-to-work (fixes deployed, monitoring required)

-- security/issue-checker