GitHub - phildougherty/llmq: Go utility for quick searches from the command line

3 min read Original article ↗

llmq

llmq is a command-line utility that interacts with the OpenAI inference service to process piped input and provide intelligent responses. It can be used to analyze logs, manage system processes, or even answer general queries using AI models like GPT-4o-mini.

Features

  • AI-Powered Command Line: Send your command line output directly to an AI model for analysis and feedback.
  • Flexible Input Handling: Works with piped input from any command-line tool.
  • Customizable: Configure default models and API keys through a YAML configuration file.

Installation

Prerequisites

  • Go: Make sure you have Go installed on your machine.
  • API Key: You'll need an API key for the OpenAI service.

Clone the Repository

git clone https://github.com/phildougherty/llmq.git
cd llmq

Build the Project

Configuration

Create a configuration file at ~/.config/llmq.yaml:

default_model: "gpt-4o-mini"
api_key: "YOUR_API_KEY"
log_level: "info"

Replace "YOUR_API_KEY" with your actual OpenAI API key.

Usage

Basic Usage

You can use llmq by piping input into it and providing a query:

echo "toast" | ./llmq --query "What is the best topping for this?"

Interesting Use Cases

Here are some creative and useful ways to integrate llmq into your Linux command-line environment:

1. Analyze System Logs for Errors

Quickly scan logs for issues:

cat /var/log/syslog | ./llmq --query "Does anything look broken in here?"

2. Summarize Git Commit History

Get a high-level summary of recent git commits:

git log --oneline | ./llmq --query "Summarize the recent changes in this project"

3. Extract Key Points from a Large Text File

Summarize a large text file to get the key points:

cat large_text_file.txt | ./llmq --query "What are the key points in this text?"

4. Suggest Improvements to a Shell Script

Get suggestions on how to improve or optimize your shell script:

cat my_script.sh | ./llmq --query "How can I improve this shell script?"

5. Identify Resource-Intensive Processes

Find processes that are consuming a lot of resources:

ps aux | ./llmq --query "Which processes are consuming the most resources?"

6. Generate a To-Do List from a Meeting Transcript

Convert a meeting transcript into actionable items:

cat meeting_transcript.txt | ./llmq --query "Generate a to-do list from this transcript"

7. Summarize System Updates

Get a summary of the packages that were updated:

sudo apt-get update | ./llmq --query "Summarize the system updates"

Increase Timeout

If you experience timeout issues, you can increase the timeout duration in the ai/client.go file:

client := &http.Client{
    Timeout: 30 * time.Second,
}

Supported Inference Service

Currently, llmq only supports OpenAI as the inference service. Ensure you have a valid API key from OpenAI to use this tool.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Reporting Issues

If you find a bug or have a feature request, please create an issue on GitHub.

License

This project is licensed under the MIT License. See the LICENSE file for details.