For ML practitioners, adding LLM calls to logic programming enables compact experiments that combine symbolic rules with generative text, useful for prototyping reasoning, data transformation, and agent scaffolding. The GitHub repository "vagos/llmpl" provides a minimal Prolog helper named pllm that exposes the predicate llm/2 to post prompts to HTTP LLM endpoints, according to the repository README on GitHub. The README shows support for any OpenAI-compatible chat/completions endpoint and documents example configuration using https://api.openai.com/v1/chat/completions with model gpt-4o-mini, as well as a local Ollama endpoint example, per the repo. The project includes installation hints (pack_install) and short usage examples that return text like "Bonjour !", and GitHub shows the repository has 1 star.
A Prolog library for interfacing with LLMs