Email Processing MCP Server

# Database Handlers ## Overview The system integrates three specialized databases, each serving a specific purpose in the email processing and analysis pipeline. ## MongoDB Handler ### Purpose - Store structured email metadata and vector embeddings - Track processing status and job information - Maintain email categories and summaries - Store emails from Inbox, Sent Items, and optionally Deleted Items ### Implementation (MongoDBHandler.py) ```python class MongoDBHandler: # Manages connection to MongoDB # Handles CRUD operations for email metadata # Tracks processing jobs and status ``` ### Key Operations - Save email metadata - Check email existence - Update processing status - Retrieve email information - Track job IDs and progress ## SQLite Handler ### Purpose - Store primary email data - Track processing status - Enable efficient filtering and retrieval - Store emails from all configured folders ### Implementation (SQLiteHandler.py) ```python class SQLiteHandler: # Manages SQLite database connection # Handles email storage and retrieval # Tracks processing status # Supports proper connection management ``` ### Key Operations - Add embeddings - Search similar content - Manage collections - Check document existence - Retrieve embedding counts ## Neo4j Handler ### Purpose - Store email relationship graphs - Enable network analysis - Track communication patterns ### Implementation (Neo4jHandler.py) ```python class Neo4jHandler: # Manages Neo4j graph database # Creates and maintains relationship graphs # Provides graph querying capabilities ``` ### Key Operations - Create email relationships - Manage vector indices - Create and maintain constraints - Execute graph queries - Close connections properly ## Integration Points ### Data Flow 1. Email metadata → MongoDB 2. Vector embeddings → ChromaDB 3. Relationship data → Neo4j ### Synchronization - Consistent Entry_IDs across databases - Status tracking in MongoDB - Job-based processing coordination ## Best Practices ### Connection Management - Proper initialization and cleanup - Connection pooling where applicable - Error handling and retries ### Data Consistency - Atomic operations when possible - Transaction support where needed - Cross-database integrity checks ### Performance - Batch operations for efficiency - Index optimization - Connection reuse ## Configuration Each handler requires specific environment variables: ### MongoDB ``` MONGODB_URI=mongodb://localhost:27017/MCP COLLECTION_NAME=outlook-emails ``` ### SQLite ``` SQLITE_DB_PATH=C:\path\to\emails.db ``` ### Email Processing ``` PROCESS_DELETED_ITEMS=true|false # Controls whether to process Deleted Items folder ``` ## Upcoming Database Integrations ### Neo4j Integration - Graph database for relationship analysis - Communication pattern visualization - Network analysis capabilities - Advanced querying for complex relationships ### ChromaDB Integration - Specialized vector database - Optimized for embedding storage and retrieval - Advanced similarity search capabilities - Efficient metadata filtering ## Error Handling - Connection failures - Query timeouts - Data validation errors - Resource cleanup - Logging and monitoring