chroma

Store embeddings and metadata for semantic document retrieval.

Updated Feb 15, 2026

Search Tags:#semantic search #embeddings #rag #vector search #chroma #document retrieval #metadata filtering

Ready to run

npx skills add https://github.com/ricable/mcai --skill chroma-ricable

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: chroma
Source: https://github.com/ricable/mcai/tree/main/.agents/skills/chroma
Command: npx skills add https://github.com/ricable/mcai --skill chroma-ricable

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Chroma eliminates manual document lookup by storing embeddings and metadata for fast semantic retrieval across collections of text, files, and knowledge bases.

Core Features & Use Cases

Open-source vector database for local and self-hosted AI applications.
Add, query, update, and delete documents while filtering by metadata for precise retrieval.
Persist collections to disk or serve them over HTTP for notebook prototypes and multi-user deployments.
Use it for RAG pipelines, semantic search, chat memory, document retrieval, and knowledge base exploration.

Quick Start

Ask for a minimal Python example that creates a Chroma collection, adds documents with metadata, and runs a similarity search.

Frequently Asked Questions about chroma

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I build semantic search with embeddings for document retrieval?▼

Semantic search with embeddings is built by storing text and metadata in a vector database for fast similarity retrieval. This eliminates manual document lookup across collections of files and knowledge bases.

Can I use vector search and metadata filtering together in a RAG workflow?▼

Vector search and metadata filtering work together in RAG workflows by adding documents with metadata and running filtered similarity searches. This allows precise document retrieval from local or self-hosted knowledge bases.

Does Chroma support local development and self-hosted production deployments?▼

Chroma supports local development by persisting collections to disk for notebook prototyping, and scales to self-hosted production deployments by serving collections over HTTP for multi-user access.

How do I manage collections for similarity search and full-text search?▼

Collection management for similarity and full-text search involves creating collections, adding documents with metadata, and updating or deleting entries. This maintains accurate knowledge bases for semantic retrieval.

What is the best way to persist embeddings for chat memory and knowledge bases?▼

The best way to persist embeddings for chat memory and knowledge bases is saving collections to disk or serving them over HTTP. This ensures data continuity across notebook prototypes and multi-user deployments.

When do I need a vector database for semantic search instead of manual lookup?▼

You need a vector database for semantic search when manual document lookup becomes too slow for large text collections. Storing embeddings and metadata enables fast similarity retrieval across RAG pipelines and knowledge bases.