Store and query embedding vectors with metadata for semantic retrieval.
npx skills add https://github.com/MarbleSodas/Mavis --skill chroma-marblesodas
Please help me install this Agent Skill. Skill: chroma Source: https://github.com/MarbleSodas/Mavis/tree/main/optional-skills/mlops/chroma Command: npx skills add https://github.com/MarbleSodas/Mavis --skill chroma-marblesodas
Chroma helps teams store document embeddings and retrieve the most relevant content for semantic search and RAG, without losing the context provided by metadata.
Use Chroma to create a local collection, add documents with metadata, and query it for semantically similar results.
High-intent search queries and answers about installing and using this skill.
To store and query document embeddings with metadata, you can create a local collection, add documents with their metadata, and run vector similarity queries. This enables semantic retrieval for RAG pipelines and document search without losing contextual attributes.
Yes, metadata filtering allows you to narrow semantic search results using attributes like source, category, page, timestamps, and custom tags. This ensures vector similarity queries return only documents matching your specific metadata constraints.
Yes, Chroma is integration-ready and works with common ML/LLM tooling such as LangChain and LlamaIndex. This allows you to build retrieval pipelines by providing a simple API for adding documents and querying embedding vectors.
You can persist and reload a vector database from disk using local-first persistence features. This supports repeatable development cycles and meets self-hosted usage requirements for saving and reloading collections during production deployments.
The best way to find relevant documents by embedding similarity is using semantic vector search. This technique retrieves the most relevant content for question answering and content discovery by comparing query embeddings against stored document vectors.
Yes, sentence-transformers is required as a dependency to generate the embedding vectors. It works alongside chromadb to enable the storage and querying of embeddings alongside metadata for similarity-based filtering across local notebooks.
jonnabioIndex and retrieve documents using semantic embeddings and metadata.
hochoa13Provides scalable vector similarity search and hybrid search with metadata filtering for RAG systems.
JhabbigBuild retrieval-augmented generation systems that ground LLM answers in external knowledge.
AnandaAnugrahHandyantoStore and search embeddings locally with a self-hosted vector database.
