https://github.com/adityanuar/clai

AI powered cli suggestion and auto complete. You don't have to remember cli commands & options, this AI tool is here to help you.
https://github.com/adityanuar/clai

Last synced: about 18 hours ago
JSON representation

AI powered cli suggestion and auto complete. You don't have to remember cli commands & options, this AI tool is here to help you.

• Host: GitHub
• URL: https://github.com/adityanuar/clai
• Owner: adityanuar
• License: apache-2.0
• Created: 2024-05-12T16:00:03.000Z (6 months ago)
• Default Branch: main
• Last Pushed: 2024-05-18T08:45:16.000Z (6 months ago)
• Last Synced: 2024-05-19T01:27:52.980Z (6 months ago)
• Language: Go
• Size: 1.24 MB
• Stars: 3
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• project-awesome - adityanuar/clai - AI powered cli suggestion and auto complete. You don't have to remember cli commands & options, this AI tool is here to help you. (Go)

README

        
# Clai 📟

Clai is an open-source CLI tool powered by AI that assists in providing CLI command suggestions using natural language. This tool utilizes open-source models and can be self-hosted, making it highly customizable and privacy-friendly. It is particularly relevant for enterprise environments where the confidentiality of internal tools is paramount. Huge thanks to [LLMWare](https://llmware.ai) for providing the tools and LLM models, making all this possible.

# Preview 📷

![Preview](img/preview.gif)

More details: 

[Youtube](https://www.youtube.com/watch?v=oqDiBObsj4o)

# Usage 🎚️

## Requirements 🛠️

Tested on:

- Golang 1.22.3

- Python 3.10.12

## Running 🏃‍♂️

### Server:

The server is used to perform inference for command suggestions from the AI model we utilize.

- It is highly recommended to use a virtual environment in Python. For example: `python -m venv [your-venv-name]`.

- go to server folder then run: `pip install -r requirements.txt`

- run server: `fastapi dev main.py`

- Your server will run on port **8000** by default. 

### Cmd:

In this folder, there is a client tool which is the core of this program. It can be used by invoking it directly from the shell. Developed using Golang, it is multi-platform and supports cross-compilation.

- go to cmd folder

- make sure your **server_url** correctly points to your server in **config.json** file

- make sure you have all dependencies ready: `go mod tidy`

- build the binary: `go build clai.go`

- use it: `./clai [your query]`