clag - Command Line AI Gem
Tired of trying to remember the exact flags to use or digging through
documentation or googling to find how to do the thing you're trying to do?
Suffer no more! Simply describe what you're trying to do and generate the
command with the help of an LLM!
Installation
Using OpenAI's GPT-4
Using Google's Gemini 1.0
-
Get an API key from Google's AI Studio at https://ai.google.dev/
-
Set your API key as GEMINI_API_KEY in your environment
-
Select Gemini as your preferred LLM by setting CLAG_LLM=gemini in your
environment
Using Anthropic's Claude 3 Opus
-
Get an API key from Anthropic at https://www.anthropic.com/
-
Set your API key as ANTHROPIC_API_KEY in your environment
-
Select Claude 3 Opus as your preferred LLM by setting CLAG_LLM=claude in
your environment
Using Groq on Mixtral
-
Get an API key from https://console.groq.com/
-
Set your API key as GROQ_API_KEY in your environment
-
Select Groq as your preferred LLM by setting CLAG_LLM=groq in your environment
Using a Local Model
-
Have a model locally from either Ollama or Llamafile with an OpenAI compatible
API
-
Have the API server running on port 8080
-
Select local as your preferred LLM by setting CLAG_LLM=local in your environment
Usage
Currently support one command: "g".
clag g "the command you'd like to generate"
Contributing
Bug reports and pull requests are welcome on Github at
https://github.com/sublayerapp/clag
Like what you see, or looking for more people working on the future of
programming with LLMs? Come join us in the Promptable Architecture
Discord