MongoDB Indexing and Query Optimization

When to use

Invoke when MongoDB queries or aggregation pipelines miss their SLO, when reviewing an inherited index set for waste or gaps, when a collection has gone hot, or when validating shard-key effectiveness against real query routing.

Do not use for: document modeling, $jsonSchema validators, or shard-key choice (use mongodb-data-model-and-migration); replica read-preference topology (use mongodb-replication-and-ha-readiness); backup/restore (use mongodb-backup-and-operational-readiness); auth/RBAC/encryption (use mongodb-security-and-data-access-hardening); ODM/repository query code (the backend implementation skill owns that — this skill tunes the engine, not application code).

Inputs

Required:

• An existing MongoDB data model and index set from mongodb-data-model-and-migration (the collections, validators, and declared access patterns this skill audits against).
• Approved performance query SLOs (latency/throughput targets per query class), or explicit confirmation they are intentionally deferred.

Optional: