ChatGPT CLI

Use this to talk with ChatGPT from your terminal.

Usage

# run app
make run

# general chatting
> hi chatbot

# for usage help
> /help

# generate an image, like a cat. uses the DALL·E model
> /image cat

# see history of conversation so far
> /history

# by default, every message is part of the same context/conversation
# this resets your context/conversation
> /reset

Context

Context is maintained for all chat inputs only. Slash commands are not captured. Use /reset to clear your context.

Prequisites

1. OpenAI API Key
2. OpenAI Org ID
3. Ruby v3+
4. Bundler v2.3+

Running Using Container

podman run --rm -e ACCESS_TOKEN=rh-35xQDkp8RUynsi6C/qU4TdZKbDkR0thisisobviouslyfake -e ORG_ID=org-UZZePthisisobviouslyfake -it quay.io/jkeam/chatgpt-cli:latest ./run.sh

# or use this image if you are on an m1 mac
# podman run --rm -e ACCESS_TOKEN=rh-35xQDkp8RUynsi6C/qU4TdZKbDkR0thisisobviouslyfake -e ORG_ID=org-UZZePthisisobviouslyfake -it quay.io/jkeam/chatgpt-cli:m1-latest ./run.sh

Running from Source

Setup

1. Clone project

   git clone [email protected]:jkeam/chatgpt-cli.git
   cd chatgpt-cli

2. Install gems

   gem install bundler
   bundle install

3. Create .env file like below, replacing with your values obtained from Prequisites:

   ACCESS_TOKEN=rh-35xQDkp8RUynsi6C/qU4TdZKbDkR0thisisobviouslyfake
   MODEL_NAME=openai-4o
   # optionally
   ORG_ID=org-UZZePthisisobviouslyfake
   URI_BASE=https://oai.hconeai.com/

Running

./run.sh
# or make run

Development

# install dev/test dependencies
bundle config set --local with test development

# install test and development dependencies
bundle install

# run tests
make test

References

1. OpenSSL with Ruby version compat