Is the indexing process slow for large libraries?

The initial index takes time to generate embeddings, but subsequent runs are incremental. It uses content hashes to skip unchanged files, making library updates very fast.

How does Zotero RAG differ from standard grep search?

Unlike grep, which finds exact keyword matches, Zotero RAG uses semantic embeddings to find passages conceptually related to your query, even if the specific words don't match.

Do I need to manage a vector database manually?

No. The skill uses ChromaDB with file-based persistence at library/.rag-index/, and all dependencies are automatically handled by the uv package manager.

What file formats does this skill support?

It indexes markdown files stored in your library. These are typically converted from PDFs or EPUBs using the included ingest scripts and must be registered in a references.bib file.

Zotero RAG Library Search

Name: Zotero RAG Library Search
Author: nealcaren

bynealcaren

•

Data Science & ML

Powers semantic search and contextual retrieval across local libraries of academic papers and markdown documents.

This skill implements a robust Retrieval-Augmented Generation (RAG) system tailored for academic researchers and sociologists using Claude Code. By indexing markdown-converted papers from a Zotero-linked library, it enables users to perform deep semantic searches that look beyond keywords to find conceptual meaning. It facilitates complex literature reviews by allowing researchers to discover thematic connections across multiple documents, find similar passages across their entire library, and retrieve the full context of specific excerpts, significantly accelerating the academic writing and synthesis process.

Key Features

01Incremental indexing of markdown files linked to BibTeX references

02Automated dependency management via uv and PEP 723 inline metadata

03Similarity discovery to find related passages across different papers

04Semantic search using sentence-transformers (all-MiniLM-L6-v2) and ChromaDB

053 GitHub stars

06Context retrieval to view surrounding text for any search hit

Use Cases

01Quickly retrieving supporting evidence and context for academic citations

02Conducting conceptual literature reviews where keyword matching is insufficient

03Identifying cross-document thematic links during qualitative data analysis

Key Features

01Incremental indexing of markdown files linked to BibTeX references

02Automated dependency management via uv and PEP 723 inline metadata

03Similarity discovery to find related passages across different papers

04Semantic search using sentence-transformers (all-MiniLM-L6-v2) and ChromaDB

053 GitHub stars

06Context retrieval to view surrounding text for any search hit

Use Cases

01Quickly retrieving supporting evidence and context for academic citations

02Conducting conceptual literature reviews where keyword matching is insufficient

03Identifying cross-document thematic links during qualitative data analysis