Consolidate duplicate API health endpoints

[Wirasm]

Problem

We currently have multiple competing health endpoints that create confusion and maintenance overhead:

Current Problematic Architecture:

• Main app: /health and /api/health (in main.py)
• Knowledge API: /api/health (in knowledge_api.py) - ROUTE COLLISION
• Projects API: /api/projects/health
• Settings API: /api/settings/health
• MCP API: /api/mcp/health
• Bug Report API: /api/bug-report/health

Issues:

1. Route collision: Two different /api/health endpoints exist
2. Confusing for clients: Frontend doesn't know which health endpoint to call
3. Duplicate logic: Migration detection logic had to be added to knowledge health because that's what frontend actually hits
4. Maintenance burden: Health logic scattered across multiple files
5. Inconsistent responses: Different endpoints return different response formats

Evidence:

Frontend logs show it's calling /api/health but hitting the knowledge module's health endpoint (returns service: 'knowledge-api') instead of the main health endpoint (should return service: 'archon-backend').

Proposed Solution

Consolidate to single canonical health endpoint:

Option 1: Single Comprehensive Health (Recommended)

• Remove all module-specific health endpoints
• Have ONE /api/health endpoint in main.py that checks:
  • Database connectivity
  • Schema migrations
  • Credential initialization
  • Critical service dependencies
• Frontend calls one endpoint, gets complete system status

Option 2: Hierarchical Health

• Keep /api/health in main.py as primary endpoint
• Move module healths to non-colliding paths: /api/knowledge/health, /api/projects/health
• Main health calls module healths and aggregates response
• Frontend still calls one canonical endpoint

Benefits

• No more route collisions
• Single source of truth for system health
• Easier to maintain health logic
• Clearer for frontend developers
• Consistent response format
• Better debugging experience

Acceptance Criteria

• [ ] Remove duplicate /api/health endpoints
• [ ] Consolidate health logic into main health endpoint
• [ ] Update frontend to call single canonical health endpoint
• [ ] Test that migration detection still works
• [ ] Ensure no breaking changes for existing health monitoring

Impact

High Priority: This architectural issue is causing real problems with our migration detection system and creates ongoing maintenance burden.