azure-cosmosdb

Azure Cosmos DB for AI, RAG, and search: vector search (vector index types, embeddings container setup, similarity / nearest-neighbor queries), full-text search (enable FTS, BM25 relevance ranking, hybrid keyword + vector), LangChain and LangGraph integration (CosmosDB checkpoint saver, async checkpointer init, MCP persistent sessions, MCP tool content and filtering, LangChain JS vectorstore, chat history, embedding model, semantic cache), and AI-agent design patterns (multi-agent routing, chat-history separation, resume from checkpoint, grounding access, background-task writes). USE FOR: store embeddings and run similarity search, choose a vector index, enable full-text/BM25 search, hybrid search, build a RAG app, persist LangGraph checkpoints, multi-agent state, MCP tools, semantic cache. DO NOT USE FOR: core modeling/partition/index/query (use cosmosdb-data-and-queries); throughput/global/monitoring/security (use cosmosdb-operations); non-AI SDK client setup (use cosmosdb-sdk).

Azure Cosmos DB performance optimization and best practices guidelines for NoSQL, partitioning, queries, and SDK usage. Use when writing, reviewing, or refactoring code that interacts with Azure Cosmos DB, designing data models, optimizing queries, or implementing high-performance database operations. USE FOR: Cosmos DB NoSQL, partition key design, RU optimization, point reads, cross-partition queries, SDK singleton, CosmosClient, container modeling, change feed, bulk operations, vector search, full-text search, hierarchical partition keys, global distribution, autoscale throughput, indexing policy. DO NOT USE FOR: PostgreSQL, MySQL, MongoDB (non-Azure), DynamoDB, Cassandra, Azure SQL, Cosmos DB for PostgreSQL (vCore), Cosmos DB for MongoDB vCore, Azure DocumentDB, general SQL databases, Redis, Elasticsearch.

Azure Cosmos DB data design and querying: document modeling (embed vs reference, denormalization, schema versioning, 2MB limit, id constraints), partition key design (high cardinality, avoid hotspots, hierarchical, synthetic keys), indexing policy (composite, range vs hash, exclude unused paths, spatial), query optimization (point reads, avoid cross-partition scans, projections, pagination, parameterization, filter order), and data-access patterns (change-feed materialized views, ranking, service layer). USE FOR: choose partition key, embed vs reference, denormalize for reads, composite index for ORDER BY, cut write RU by excluding paths, point read by id and partition key, paginate with continuation tokens, avoid cross-partition queries, materialized views. DO NOT USE FOR: SDK client code (use cosmosdb-sdk); throughput, global distribution, monitoring, security (use cosmosdb-operations); vector or full-text search and LangChain/LangGraph agents (use cosmosdb-ai-and-search).

Azure Cosmos DB operations, scaling, and security: throughput provisioning (autoscale vs manual, serverless for dev, RU estimation, bursts), global distribution (multi-region writes, read regions, failover, consistency levels, conflict resolution, zone redundancy), monitoring and diagnostics (RU consumption, throttling/429 alerts, P99 latency, Azure Monitor, diagnostic logs), and security (managed identity, disable key auth, RBAC least privilege, network restriction, continuous backup / point-in-time restore). USE FOR: provision RU/s, autoscale, serverless, multi-region writes, region failover, choose consistency level, alert on throttling, track P99 latency, push metrics to Azure Monitor, authenticate with managed identity, lock down to a VNet, RBAC, restore. DO NOT USE FOR: modeling, partition keys, indexing, queries (use cosmosdb-data-and-queries); SDK client code (use cosmosdb-sdk); vector/full-text search and AI agents (use cosmosdb-ai-and-search).

Azure Cosmos DB SDK and developer tooling for .NET, Java, Python, Go, and Spring Boot: singleton CosmosClient, async APIs, Direct vs Gateway connection mode, retry/429 handling, preferred and excluded regions, availability strategy, circuit breaker, diagnostics, enum serialization, ETag optimistic concurrency, conditional create, patch counter increment, continuation-token guards, per-call request options, content response on write, Spring Data annotations and repositories, NuGet package setup, Python async deps, emulator SSL, and local development / VS Code tooling. USE FOR: reuse CosmosClient as singleton, Direct mode, handle 429 retries, preferred regions, ETag concurrency, atomic patch increment, Spring Boot config, .NET package setup, run the emulator locally, browse data in VS Code. DO NOT USE FOR: modeling/partition/index/query (use cosmosdb-data-and-queries); throughput, global, monitoring, security (use cosmosdb-operations); LangChain/LangGraph/vector/full-text search (use cosmosdb-ai-and-search).