azure-cosmosdb-performance-investigator

Azure Cosmos DB Performance Investigator

Purpose

Investigate Azure Cosmos DB performance pathologies with evidence-first profiling instead of lazy “add more RUs” advice.

This skill is for deep performance work across:

• RU inefficiency and unexpected request-charge spikes,
• query latency and scan-heavy query behavior,
• hot partitions and partition-key skew,
• throttling, retry inflation, and client-perceived latency,
• indexing gaps and poor query/index alignment,
• container, partition, and workload-level profiling,
• diagnostic-log and metrics-backed remediation planning.

When to use

Use this skill when the user asks for:

• slow Azure Cosmos DB queries or workload latency,
• high RU cost or suspicious request-charge behavior,
• 429 throttling analysis,
• hot partition or partition-skew investigation,
• indexing or query-performance tuning,
• a step-by-step Cosmos DB profiling plan.

Do not use this skill as a substitute for:

• initial data-model design when the main problem is greenfield schema modeling,
• pure account/platform governance review when performance is incidental,
• generic application debugging unrelated to Cosmos DB workload behavior,
• vector-search-specific Mongo vCore tuning unless the user explicitly asks for that API surface.

Lean operating rules

• Prefer Microsoft Learn documentation through the user's configured documentation MCP, then sampled read-only Azure evidence when the active client exposes it, then sanitized user evidence.
• Separate confirmed facts from inference. If state was not queried or shown, say so.
• Challenge throughput-first fixes that ignore partition skew, query scans, indexing, or client retry inflation.
• Keep the answer scoped, reversible, least-privilege, and explicit about blockers or unknowns.

References

Load these only when needed:

• Operations guide — use for service-specific pitfalls, design rules, verification targets, and pushback criteria.
• MCP and evidence path — use when choosing documentation-based evidence, sampled read-only Azure evidence, or sanitized user evidence.
• Safety checklist — use for evidence labels, risk gates, mutation boundaries, approval rules, and credential boundaries.
• Workflow and output contract — use when executing the full investigation, applying stress checks, or formatting the final answer.
• Data profiling playbook — use when you need the detailed step-by-step profiling sequence.
• Official sources — use when you need the detailed Microsoft documentation list or source notes.