Package: llm
A unified abstraction over large-language-model providers. Callers code against a single interface and swap providers via configuration — the same prompt code works with Anthropic, Gemini, OpenAI, OpenRouter, or Vertex AI.
Repository: https://github.com/dracory/llm
Installation
go get github.com/dracory/llm
Features
- Single provider interface covering chat completions.
- Providers: Anthropic, Google Gemini, OpenAI, OpenRouter, Vertex AI.
- Provider selection at runtime — each is enabled and configured via its
*_API_USED/*_API_KEY/*_DEFAULT_MODELenvironment keys.
Where It Is Used
The app exposes an LLM factory (app.Llm) which the external admin modules
(blogadmin, shopadmin) consume through adapters for AI-assisted content generation.
See Environment Variables for the provider
configuration keys.