Cosmosdb
Azure Cosmos DB multi-model database with global distribution. Use for Azure.From its SKILL.md
npx -y skills add G1Joshi/Agent-Skills --skill cosmosdbAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 12 stars12 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.8 KB, 426 tokens by cl100k_base, as published. Nobody here has run it
Azure Cosmos DB
Cosmos DB is Azure's planetary-scale database. It supports multiple APIs: NoSQL (Core/JSON), MongoDB, PostgreSQL, Cassandra, Gremlin (Graph), and Table.
When to Use
- Global Distribution: Replicate data to any Azure region with a click.
- Multi-Model: If you need Mongo or Postgres APIs but want "PaaS" management.
- Low Latency: Guaranteed <10ms read/write latency at the 99th percentile.
Quick Start (NoSQL API)
Container container = database.GetContainer("Items");
Item item = new Item
{
Id = "1",
Category = "Personal",
Name = "Groceries"
};
await container.CreateItemAsync(item, new PartitionKey(item.Category));
Core Concepts
Request Units (RUs)
The currency of Cosmos DB. Use RUs to pay for throughput. 1 RU ≈ reading a 1KB doc.
Partition Key
Crucial. Determines how data is distributed. A bad partition key ("Date") creates "Hot Partitions" (bottlenecks). A good key ("UserId") distributes load evenly.
Consistency Levels
Offers 5 levels: Strong, Bounded Staleness, Session (Default), Consistent Prefix, Eventual. Trade off consistency for availability/latency.
Best Practices (2025)
Do:
- Use the NoSQL API: It is the native API with the most features.
- Use Hierarchical Partition Keys (2025): Supports up to 3 keys for better data distribution (TenantId -> UserId -> DeviceId).
- Use Analytical Store (Synapse Link): Run heavy analytics (BI) on your operational data without impacting performance.
Don't:
- Don't ignore RU consumption: Monitor it. Queries without a partition key ("Cross-partition queries") are expensive.
References
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.