Codepit The toolbox

A Qdrant architect's handbook for Claude

The skills help Claude make the Qdrant decisions that decide whether vector search works well, like quantization, sharding, tenant isolation and hybrid search. They answer when and why, then point to the exact Qdrant docs page.

Scan checks all of qdrant/skills on your computer, before anything installs.

Codepit runs on your computer. Open this page there to scan it before installing.

An open manual with a honey ribbon

What you get

14 skills, each routing Claude from your symptom to the right guidance.

  • qdrant-scaling
  • qdrant-sizing
  • qdrant-performance-optimization
  • qdrant-search-quality
  • qdrant-multitenancy
  • qdrant-model-migration
  • qdrant-version-upgrade
  • qdrant-monitoring
  • qdrant-deployment-options
  • qdrant-clients-sdk

When it kicks in

  • Slow search
  • Irrelevant results
  • Sizing a cluster
  • Scaling up or out
  • Many tenants
  • Switching embedding models
  • Version upgrades

Claude matches what you describe to a skill and reads that guidance before it answers.

Try saying

>I have 50M vectors on a single node and search is slow, should I add more nodes?

Claude opens the scaling guidance for your symptom and checks whether scaling the node up fits before suggesting sharding across more nodes.

Worth knowing

The skills are under active development, so their content and structure may change between versions as Qdrant evolves.

Read the author's README README.md