AIRA-43: Incident Manager Core Service

[learn-design-develop]

👤 User Story

As the core system, I need an incident manager to orchestrate webhook processing, deduplication, state management, and persistence.

🎯 Rationale

Single-worker orchestrator for Phase 1.

✅ Acceptance Criteria

• IncidentManager methods: create_incident(), update_incident(), get_incident_by_id(), get_incidents_by_status()
• Use SQLAlchemy session with rollback and retry on transient errors
• Invoke DedupService and StateMachine
• JSON logging (component, incident_id, duration_ms, status, correlation_id)
• Handle DB failures, Redis timeouts, unexpected exceptions
• Unit & integration tests

📋 Metadata

• Status: MVP

• Category: Core Workflow

• Week: Week 4

• Complexity: Medium

• Critical Path: Yes

• Dependencies: AIRA-40, AIRA-41

---

Original Ticket: #43
Phase 1 MVP Tracking Issue