Automate Milvus collection creation and vector index management.
npx skills add https://github.com/zilliztech/milvus-marketplace --skill indexing
Please help me install this Agent Skill. Skill: indexing Source: https://github.com/zilliztech/milvus-marketplace/tree/main/plugins/core/skills/indexing Command: npx skills add https://github.com/zilliztech/milvus-marketplace --skill indexing
Managing Milvus vectors requires careful coordination of collections, schemas, and index configurations. This Skill automates the setup process to ensure efficient vector search and scalable storage.
Define your Milvus schema and create an embedding-enabled collection named my_collection. Configure an HNSW index for a small dataset, then run a vector search with your query embedding.
High-intent search queries and answers about installing and using this skill.
Creating a Milvus collection involves defining a schema with id, text, and embedding fields, then configuring an appropriate vector index. This Skill automates that setup—choose HNSW for high-precision small datasets, IVF_FLAT for mid-sized data, or IVF_PQ for very large datasets to balance latency and memory usage.
HNSW delivers high precision for small datasets but uses more memory; IVF_FLAT suits mid-sized data with moderate latency; IVF_PQ compresses vectors for very large datasets, reducing memory at the cost of precision. This Skill helps you choose and tune the right index type for your data scale and latency requirements.
Yes, this Skill supports scalar indexing alongside vector indexes for collections. Scalar indexes optimize filtering on metadata fields, enabling hybrid search workflows like product search or document retrieval that combine vector similarity with attribute filtering.
This Skill automates collection and schema creation with structured fields, then applies the appropriate vector index and optional scalar indexing. It handles the full setup for retrieval workflows, letting you focus on query embeddings and search logic.
This Skill guides index selection and tuning across dataset sizes—HNSW for small, IVF_FLAT for medium, IVF_PQ for large. It handles configuration to balance latency and memory, so you avoid manual parameter guessing based on your data scale.
Yes, this Skill supports creating collections with optional partitions to improve query performance. Partitioning reduces search scope, making vector indexing more efficient especially for very large or time-series datasets.
patricio0312revAutomate document embedding pipelines for semantic search.
JhabbigImplement similarity search patterns for vector databases and ranked retrieval.
tranhieuttImplement vector database workflows for semantic search retrieval.
davpatel605-beepBuild vector similarity search systems with Qdrant for RAG applications.
