🌐 Memory Orders API

📍 Location

• Service: services/api_memory_orders/
• Pure core: vitruvyan_core/core/governance/memory_orders/

🚪 Base URL / Port

• Config default: MEMORY_API_PORT=8016 (services/api_memory_orders/config.py)

Deploy note: other documents may refer to a different “fleet” port. Treat docker-compose mapping as the source of truth.

🧭 Endpoints (implemented)

Defined in services/api_memory_orders/api/routes.py:

• GET / — service info
• GET /health — simple health check
• GET /health/rag — full dependency + coherence health (returns 503 when degraded/critical)
• POST /coherence — drift analysis between Postgres and Qdrant
• POST /sync — sync planning (planning only, no execution)
• GET /metrics — Prometheus (defined in services/api_memory_orders/main.py)

🧾 Request/Response models

See services/api_memory_orders/models/schemas.py.

📉 POST /coherence

Request (CoherenceRequest):

• table: str = "entities"
• collection: str | null
• embedded_column: str = "embedded"

Response (CoherenceResponse):

• status: healthy|warning|critical
• drift_percentage: float
• recommendation: str
• pg_count: int
• qdrant_count: int
• drift_absolute: int
• timestamp: str
• table: str
• collection: str

🔁 POST /sync

Request (SyncRequest):

• mode: incremental|full (default incremental)
• table: str = "entities"
• collection: str | null
• limit: int = 1000

Response (SyncResponse):

• operations: SyncOperationResponse[]
• estimated_duration_s: float
• mode: str
• total_operations: int

⚙️ Env vars (service-level)

Loaded in services/api_memory_orders/config.py:

• Service: MEMORY_API_PORT, LOG_LEVEL
• Postgres: POSTGRES_HOST, POSTGRES_PORT, POSTGRES_DB, POSTGRES_USER, POSTGRES_PASSWORD
• Qdrant: QDRANT_HOST, QDRANT_PORT, QDRANT_COLLECTION
• Redis: REDIS_HOST, REDIS_PORT, REDIS_DB, REDIS_URL
• External checks: EMBEDDING_API_URL, BABEL_GARDENS_URL
• Coherence thresholds: COHERENCE_THRESHOLD_HEALTHY, COHERENCE_THRESHOLD_WARNING
• Audit: MEMORY_AUDIT_TABLE
• Flags: ENABLE_AUTO_SYNC, ENABLE_PROMETHEUS