v0.5.1

Release Notes - ask-sh v0.5.1

This release introduces a significant new feature that expands the compatibility of ask-sh with various LLM providers and enhances flexibility in choosing your preferred AI backend.

Features

• OpenAI Compatible API Support: You can now use ask-sh with APIs that are compatible with the OpenAI API, such as:

  • Ollama: Run local LLMs like Llama 3 or DeepSeek models directly on your machine.
  • DeepSeek API: Utilize the DeepSeek API for alternative models and potentially different pricing.

  To use this feature, set the ASK_SH_OPENAI_BASE_URL environment variable to the base URL of your chosen OpenAI-compatible API endpoint.

  Examples:

  • Ollama: export ASK_SH_OPENAI_BASE_URL="http://localhost:11434/v1"
  • DeepSeek: export ASK_SH_OPENAI_BASE_URL="https://api.deepseek.com"

  Refer to the updated README for detailed instructions and configuration options.

Documentation Updates

• Expanded LLM Provider Guide: The README has been updated to include comprehensive information and examples for utilizing OpenAI-compatible APIs. This includes:
  • Clear instructions on how to configure ASK_SH_OPENAI_BASE_URL.
  • Examples for popular options like Ollama and DeepSeek.
  • Updated list of supported LLM providers to explicitly mention OpenAI-compatible APIs.

This release empowers you to leverage a wider range of LLM options with ask-sh, providing greater control and customization over your AI command-line assistant experience.