Skip to content

Latest commit

 

History

History
235 lines (152 loc) · 6.59 KB

setup.md

File metadata and controls

235 lines (152 loc) · 6.59 KB

Setup Guide

This guide explains how to install and set up wish-sh and wish-knowledge-loader.

Prerequisites

  • Python 3.13 or higher
  • OpenAI API key (available from the OpenAI website)

Setup Overview

This guide will walk you through the process of setting up the wish ecosystem components in the correct order. Follow these steps to ensure proper functionality:

  1. Environment Setup: Configure the necessary environment variables for all components
  2. Start the API Server: Launch the wish-log-analysis-api server using make run-api
  3. Install and Configure wish-sh: Install the wish-sh package and set up required configurations
  4. Install wish-knowledge-loader (Optional): Set up the knowledge loader if you need enhanced functionality

The wish-sh command requires the wish-log-analysis-api server to be running, as it relies on the API for analyzing command execution logs. This setup ensures that all components work together seamlessly.

The following sections provide detailed instructions for each step of the setup process.

wish-log-analysis-api

Environment Variables

wish-log-analysis-api automatically reads environment variables from the ~/.wish/env file. Set the following environment variables:

# OpenAI API settings
OPENAI_API_KEY=your-api-key-here
OPENAI_MODEL=gpt-4o

# API settings
WISH_API_BASE_URL=http://localhost:3000

See the "Using env File" section for instructions on setting up the environment variable file.

Note: The previous method (setting environment variables using export commands) does not affect the API server running in a container. Environment variables are automatically loaded from the ~/.wish/env file and passed to the container.

Starting the API Server

To start the wish-log-analysis-api server locally, run the following command:

make run-api

This will start a local development server and make the API endpoints available.

wish-sh

Installation

Install wish-sh using the following command:

pip install wish-sh

Environment Variables

To use wish-sh, you need to set the following environment variables:

# Required: Set your OpenAI API key
export OPENAI_API_KEY=your-api-key-here

# Optional: Set the OpenAI model to use (default: gpt-4o)
export OPENAI_MODEL=gpt-4o

# Optional: Set the wish home directory (default: ~/.wish)
export WISH_HOME=~/.wish

# Optional: Set a custom env File path (default: $WISH_HOME/env)
# This is used by all wish packages
export WISH_ENV_FILE=/path/to/your/custom/.env

You can add these environment variables to your shell configuration file (like .bashrc or .zshrc) to have them automatically set when you open a terminal.

Using env File

Instead of setting environment variables directly, you can use a .env file. The default location for this file is $WISH_HOME/env. Here's how to set it up:

  1. Make sure the $WISH_HOME directory exists:
mkdir -p ~/.wish
  1. Copy the example .env file to $WISH_HOME/env:
cp .env.example ~/.wish/env
  1. Edit the $WISH_HOME/env file to set your actual API keys and other settings:
nano ~/.wish/env

The .env file should contain settings like:

# Wish home directory
WISH_HOME=~/.wish

# OpenAI API settings
OPENAI_API_KEY=your-api-key-here
OPENAI_MODEL=gpt-4o
# Required for wish-knowledge-loader
OPENAI_EMBEDDING_MODEL=text-embedding-3-small

# LangSmith settings (optional)
LANGCHAIN_TRACING_V2=true
LANGCHAIN_ENDPOINT=https://api.smith.langchain.com
LANGCHAIN_API_KEY=your-langsmith-api-key-here
LANGCHAIN_PROJECT=wish

Make sure to replace your-api-key-here with your actual OpenAI API key, and if you're using LangSmith, replace your-langsmith-api-key-here with your actual LangSmith API key.

Verifying Installation

After installation and environment variable setup, you can start wish-sh with:

wish

Note for macOS users: macOS includes a built-in wish command as part of the Tcl/Tk package. To avoid conflicts, you can use the alternative command:

wish-sh

Both commands provide identical functionality.

You can also specify a custom env File path using the --env-file option:

wish --env-file /path/to/your/custom/.env
# or on macOS
wish-sh --env-file /path/to/your/custom/.env

wish-knowledge-loader

Installation

Install wish-knowledge-loader using the following command:

pip install wish-knowledge-loader

Environment Variables

wish-knowledge-loader uses the same environment variables as wish-sh (OPENAI_API_KEY is required, while OPENAI_MODEL and WISH_HOME are optional).

Additionally, when using wish-knowledge-loader, you need to set the following environment variable:

# Required for wish-knowledge-loader: Set the OpenAI embedding model
export OPENAI_EMBEDDING_MODEL=text-embedding-3-small

This variable is specifically required for the knowledge loader functionality, as it needs to generate embeddings for the documents.

Verifying Installation

After installation, you can verify that wish-knowledge-loader is working correctly by running:

wish-knowledge-loader --help

You should see a help message similar to:

Usage: wish-knowledge-loader [OPTIONS]

Options:
  --repo-url TEXT    GitHub repository URL to clone
  --glob TEXT        Glob pattern for files to include
  --title TEXT       Title for the knowledge base
  --help             Show this message and exit.

Troubleshooting

Common Issues

  1. API Key Error

    Error: OpenAI API key not found

    Solution: Ensure that the OPENAI_API_KEY environment variable is correctly set.

  2. Model Error

    Error: The model 'xxx' does not exist

    Solution: Ensure that the OPENAI_MODEL environment variable is set to a valid OpenAI model name.

  3. Python Version Error

    Error: Python 3.13 or higher is required

    Solution: Ensure that Python 3.13 or higher is installed.

  4. Embedding Model Error

    Error: You are not allowed to generate embeddings from this model

    Solution: Ensure that the OPENAI_EMBEDDING_MODEL environment variable is set to a valid OpenAI embedding model (e.g., text-embedding-3-small) when using wish-knowledge-loader.

Next Steps

Now that you have installed and set up wish-sh and wish-knowledge-loader, you can: