wish-knowledge-loader 0.3.0

Knowledge loader for wish

wish-knowledge-loader

A CLI tool for loading knowledge bases into wish.

Overview

wish-knowledge-loader is a command-line tool that clones GitHub repositories, extracts content from specified files, and stores them in a vector database for use with wish.

Installation

# Install from the repository
cd wish-knowledge-loader
uv sync --dev

cp .env.example .env
vim .env  # Set the OpenAI API key

Usage

# Basic usage
wish-knowledge-loader --repo-url https://github.com/username/repo --glob "**/*.md" --title "Knowledge Base Title"

Options

• --repo-url: GitHub repository URL to clone
• --glob: Glob pattern for files to include (e.g., "**/*.md" for all Markdown files)
• --title: Title for the knowledge base

Environment Variables

The following environment variables can be set in a .env file:

• OPENAI_API_KEY: Your OpenAI API key (required)
• OPENAI_MODEL: OpenAI embedding model to use (default: "text-embedding-3-small")
• WISH_HOME: Path to the wish home directory (default: "~/.wish")

Development

# Run tests
uv run pytest

# Run linting
uv run ruff check .

Verifying Search Functionality

After creating a knowledge base, you can verify the search functionality using the provided script:

# Set the OpenAI API key
export OPENAI_API_KEY="your-api-key"

# Run the search script
python scripts/search_knowledge.py "Knowledge Base Title" "your search query"

The script will:

1. Load the vector store for the specified knowledge base
2. Search for documents similar to your query
3. Display the top 4 results with their content and metadata

You can adjust the number of results by modifying the k parameter in the script.