aimi

function

1. catgirl
2. Chat AI developed based on OpenAI/gemini/bing/plugin
3. Support qq individual or group chat permission
4. mathematical expression(LaTex) / html converted to pictures
5. Extension of robots via plugins is supported
6. Translation and execution of Functions from Natural Language to programming Language

environmental requirement

1. Python version >= 3.9
2. Get a OneBot server ready, like Shamrock on https://yuyue-amatsuki.github.io/OpenShamrock
3. cp ./run/setting.yml.example ./run/setting.yml
4. An OpenAI account for your own use.

If you want to use text-to-image, you'll need to add these

1. pandoc
2. wkhtmltopdf

running

configuring dependencies

1. install dependent package, if you use python3.9, then use:

pip3.9 install -r ./run/requirements.txt

or use:

cd run && bash -x build_environment.sh

If you want to use mathematical expression(LaTex) / html to turn images and you are a class ubuntu system

apt install -y pandoc wkhtmltopdf

running as web

1. Make sure you have nodejs installed

apt-get install nodejs

2. Installing web dependencies

cd ./app/app_web/
npm i

3. Starting the Web server. When the web server has started successfully, open a url like http://localhost:2300 to view the front-end page, depending on the port on which the logs were run

npm run dev

4. run bot

cd ./ && python3.9 main.py > ./run/aimi.log

5. Fill in the api-key for chatgpt's api in the web page, Start a chat.

running as a OneBot QQ server

1. run Shamrock, go-cqhttp or another OneBot server

set setting:

qq:
  type: go-cqhttp
  port: '5701' # Local listening port
  post_host: 'localhost' # IP/domain of the QQ message parsing HTTP server
  post_port: '5700' # Port of the QQ message parsing server
  uid: 123 # QQ number used for logging into the QQ message parsing server, which is the bot's own QQ number.
  master_uid: 1234 # Administrator's QQ number
  response_user_ids: # List of QQ numbers that the bot can reply to
    - 1234
  response_group_ids: # List of group IDs that the bot can reply to
    - 123456789
  manage: 
    reply_time_limit_s: 3600 # Time interval limit for each message
    protect_bot_ids: # List of users who need speech restrictions (requires administrator permissions)
      -  12345678

2. run OneBot server, Please make sure that he can receive qq messages. You need to solve login and authentication issues, see the official user feedback for details.

3. run bot

cd ./ && python3.9 main.py > ./run/aimi.log

If you set up correctly, then you can see qq messages, as well as the administrator qq see aimi online.

Have a good time!

How to get your OpenAI access token:

(For development communication only)

https://chat.openai.com/api/auth/session

save to ./run/setting.yml

openai:
  access_token: # your access_token

6. express gratitude

https://github.com/acheong08/ChatGPT

https://github.com/Mrs4s/go-cqhttp

https://github.com/LSTM-Kirigaya/go-cqhttp-python-server

https://github.com/vercel/examples

https://github.com/mckaywrigley/chatbot-ui

https://yuyue-amatsuki.github.io/OpenShamrock/

7. release claims

Development and communication are welcome, final interpretation and copyright are reserved, based on the MIT agreement.