Skip to content

fa-ina-tic/memvid-rag

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Memvid RAG

Agent Skills for local RAG (Retrieval-Augmented Generation) using memvid-sdk.

These skills follow the Agent Skills specification so they can be used by any skills-compatible agent, including Claude Code and Codex CLI.

Installation

Marketplace (Recommended)

Install via Claude Code plugin marketplace:

/plugin marketplace add fa-ina-tic/memvid-rag
/plugin install memvid-rag@memvid-rag-skills

Claude Code (Manual)

Place the contents of this repository in a /.claude folder at your project root. See the Claude Skills documentation for details.

Codex CLI

Copy the skills/ directory to ~/.codex/skills following the Agent Skills specification.

Requirements

  • Python 3.8+
  • memvid-sdk: Install with pip install memvid-sdk
  • API Key (for cloud embeddings): Set one of the following environment variables:
    • COHERE_API_KEY - Cohere (default)
    • OPENAI_API_KEY - OpenAI
    • VOYAGE_API_KEY - Voyage AI
    • NVIDIA_API_KEY - NVIDIA

Skills

This plugin enables semantic search over PDF documents:

Command Description
/memvid-rag:create Create a new knowledge.mv2 index file
/memvid-rag:index Index PDF documents with vector embeddings
/memvid-rag:search Search indexed documents semantically
/memvid-rag:status Show RAG system statistics

Usage

1. Create Index

Initialize a new knowledge base:

/memvid-rag:create

2. Index Documents

Add PDF documents to the knowledge base:

/memvid-rag:index ./documents/research-paper.pdf

3. Search

Query your indexed documents:

/memvid-rag:search What are the key findings?
/memvid-rag:search How does the algorithm work? --k=10 --mode=sem

Search Options

  • --k=<number> - Maximum results (default: 5)
  • --mode=<lex|sem|hybrid> - Search mode (default: hybrid)
  • --snippet_chars=<number> - Max characters per snippet (default: 240)
  • --min_relevancy=<float> - Minimum relevancy threshold
  • --adaptive=<true|false> - Enable adaptive result count

Data Storage

All indexed data is stored locally in a knowledge.mv2 file in your working directory.

License

MIT

About

local, multimodal rag agent for claude code

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages