Microsoft Azure CosmosDB
Azure Cosmos DB plugin
工作原理
The official Azure Cosmos DB plugin bundles a curated set of agent skills covering data modeling, partition key design, query optimization, SDK best practices (singleton client, bulk operations, direct mode), indexing strategies (composite, spatial, exclude unused), vector search (DiskANN, flat, quantized flat), full-text search (BM25, hybrid queries), global distribution (multi-region writes, conflict resolution), security (Entra ID, managed identity, RBAC, network restrictions), throughput and scaling (autoscale, serverless), monitoring and diagnostics, and design patterns (change feed, transactional batch). It gives your agent the knowledge to generate reliable Cosmos DB code and apply best practices end to end.
Choose an orders partition key
Compare partition keys against order access patterns and expected growth to recommend a model that avoids hot partitions.
Reduce database query costs
Turn a costly query into a prioritized set of query and indexing changes with a plan to measure the improvement.
Design product similarity search
Create a vector search design for a product catalog with an index recommendation and queries that respect catalog filters.