Define systemic pattern detection — collective signals across entities

[koad]

Monitor for patterns affecting multiple entities, not just individual stalls.

Pattern: Multiple entities showing related signals that suggest a systemic issue:

• All build-dependent entities stalled after Vulcan deployment
• Multiple reviews backed up (Veritas gate bottleneck)
• Several entities waiting on same external API/service
• Contributor churn in multiple entities (team reorganization signal)

What to monitor:

• Concurrency: >2 entities stalled in same 48h window (not independent incidents)
• Pipeline cascade: failure in early stage (Vulcan) blocks multiple downstream entities
• Common blocker: >2 entities with same blocked-by label or cross-linked issue
• Shared resource contention: multiple entities modifying same file/dependency
• Team signals: multiple entities losing commits from same author (turnover? scope shift?)

Tier: WARNING (multi-entity pattern), escalate to Juno for remediation

Action: File issue on koad/juno with cross-repo references when systemic pattern detected

Related: alert-routing.md, fourty4-watch-setup.md, GOVERNANCE.md

Shifts from "entity X stalled" to "why is the system stalled" — enables team-level thinking.

[koad]

Consolidated into #15