weaviate-claude-skills

🥈Silver

Connect Claude.ai to local Weaviate vector databases for managing collections, ingesting data, and querying with RAG. Ideal for operations teams using Docker to enhance document search and semantic search capabilities.

3960Updated 1mo ago

Intermediate30min to implementautomation

Saves ~90 min per use

Quick InstallView Source

git clone https://github.com/saskinosie/weaviate-claude-skills.git

Works with:

Claude

Overview

About This Skill

How to Use

["Install the Weaviate-CLAUDE-Skills plugin in Claude.ai via the plugin manager. Ensure Docker is running locally and Weaviate is accessible on port 8080.","Use the 'weaviate create collection' command in the Claude.ai terminal to generate a new collection with the specified dimensions. Verify the collection exists in Weaviate.","Run the 'weaviate ingest data' command with the directory path containing your documents. Monitor the ingestion progress and success rate in the terminal.","Execute the 'weaviate query RAG' command with your search query to retrieve relevant results. Refine the query or collection parameters based on the output quality.","Tip: For better results, use the 'sentence-transformers' model for embeddings and split documents into chunks of 500 tokens each. Adjust the RAG parameters for more precise retrieval."]

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/saskinosie/weaviate-claude-skills

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Connect Claude.ai to your local Weaviate vector database running in Docker. Create a new collection named [COLLECTION_NAME] with [NUMBER_OF_DIMENSIONS] dimensions. Ingest the documents from [DIRECTORY_PATH] into this collection. Then, perform a semantic search for the query '[SEARCH_QUERY]' using RAG to retrieve relevant results.

Example Output

The Weaviate Docker container is running locally on port 8080, and the Weaviave-CLAUDE-Skills plugin is installed in Claude.ai. A new collection called 'technical_manuals' is created with 768 dimensions to support semantic search. The documents from the directory '/documents/manuals' are ingested into this collection, including 12 manuals in PDF and Markdown formats such as 'Docker_Quick_Start_Guide.pdf' and 'Kubernetes_Troubleshooting.md'.

The ingestion process takes 4 minutes and 12 seconds, with a success rate of 100%. The manuals are split into chunks of 500 tokens each, and the embeddings are generated using the 'sentence-transformers' model. The embeddings are stored in the 'technical_manuals' collection, along with the metadata such as the document name, the chunk number, and the token count.

A semantic search is performed for the query 'How to configure Docker networking in Kubernetes'. The RAG (Retrieval-Augmented Generation) process retrieves the top 3 relevant chunks from the 'technical_manuals' collection. The chunks include:

1. 'Docker_Networking_Configuration_Guide.pdf' - Chunk 3: 'Configure the Docker networking settings in the Kubernetes cluster...'
2. 'Kubernetes_Troubleshooting.md' - Chunk 7: 'Check the Docker networking configuration in the Kubernetes pods...'
3. 'Advanced_Docker_Deployment.pdf' - Chunk 2: 'Set the Docker networking parameters in the Kubernetes YAML files...'

The retrieved chunks are combined into a single response using the RAG process. The response includes:

'To configure Docker networking in Kubernetes, follow these steps:

1. Configure the Docker networking settings in the Kubernetes cluster using the 'Docker_Networking_Configuration_Guide.pdf'.
2. Set the Docker networking parameters in the Kubernetes YAML files as described in the 'Advanced_Docker_Deployment.pdf'.
3. Check the Docker networking configuration in the Kubernetes pods using the troubleshooting steps in the 'Kubernetes_Troubleshooting.md'.'

The semantic search process takes 23 seconds, with a success rate of 95%. The retrieved chunks are relevant and provide actionable insights for managing the technical manuals in the local Weaviate vector database.