Skip to content
Get packs

Chroma Embedding Database

Lets the agent build and query a local vector database for semantic search and RAG.

Source Download skill
Repository
Orchestra-Research/AI-Research-SKILLs
Stars
14,195
Licence
MIT
Files
2 files

What it does

When installed, the agent knows how to set up Chroma, an open-source database that stores text and its numerical 'embeddings' so it can find documents by meaning rather than exact keywords. It can write code that adds documents with metadata, runs similarity searches with metadata filters, updates or deletes entries, saves data to disk, and wires it into frameworks such as LangChain for retrieval-augmented generation (feeding relevant documents to an AI model).

Great for

  • Developers prototyping a RAG or semantic search feature in Python or JavaScript
  • Teams that want a self-hosted, open-source vector store instead of a managed cloud service
  • People building a document search tool over their own files

Skip it if

  • Large production systems needing managed, auto-scaling infrastructure, where the skill itself points to alternatives like Pinecone or Qdrant
  • Projects that must not send text to third-party embedding APIs, since the OpenAI and HuggingFace examples transmit data externally and the examples use placeholder API keys written in code

Try asking

  • “Create a Chroma collection from my folder of markdown notes and let me search it by meaning”
  • “Set up a persistent Chroma store on disk and add these product descriptions with category metadata”
  • “Filter my Chroma query results to only documents where source is 'pdf' and page is greater than 10”
  • “Wire Chroma into a LangChain retrieval chain over my company FAQ documents”
Step-by-step guide: how to use Chroma Embedding DatabaseInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 2 original files, the MIT licence and an attribution note.

    chroma.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip chroma.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip chroma.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Use PersistentClient with a fixed path so embeddings survive restarts instead of rebuilding them each session
  • Keep API keys in environment variables rather than pasting them into code, as the examples show literal placeholders
  • Check the included integration reference and the skill's dependency list before installing, since the default embedding model is downloaded on first use

What's inside

  • references/integration.md
  • SKILL.md

Redistributed unmodified from Orchestra-Research/AI-Research-SKILLs under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.