advanced-askai

This package allows you to have a chat conversation with AI right from your terminal.

Results are streamed to the console.

It has api bindings to allow programatic access and can be used for other tools.

The api key is stored using the secure keyring package. Right now we only support OpenAI's chatbot, but in the future I hope to include support for anthropics.

The only stable api in this package is located at advanced_askai.api

Binding to any other module in this library is subject to break.

Help

usage: Ask OpenAI for help with code

positional arguments:
  prompt                Prompt to ask OpenAI

options:
  -h, --help            show this help message and exit
  --input-file INPUT_FILE
                        Input file containing prompts
  --json                Print response as json
  --set-key SET_KEY     Set OpenAI key
  --output OUTPUT       Output file
  --advanced            bleeding edge model: gpt-4o
  --model MODEL
  --verbose
  --no-stream
  --assistant-prompt ASSISTANT_PROMPT
  --assistant-prompt-file ASSISTANT_PROMPT_FILE
                        File containing assistant prompt
  --max-tokens MAX_TOKENS
                        Max tokens to return
  --code                Code mode: enables aider mode
  --check               Sends the response back to the chatbot for a second opinion

See advanced_askai.api.*

Develope

To develop software, run . ./activate.sh

Windows

This environment requires you to use git-bash.

Linting

Run ./lint.sh to find linting errors using pylint, flake8 and mypy.

Releases

• 1.0.3 - Treat interactive exit case-insensitively.
• 1.0.2 - Fixes openai dependency break: openai>=1.3.7,<=1.55.3
• 1.0.1 - If the max tokens goes negative, just set it to 16000.