llm-pipeline 0.2.0

Reusable node payloads for LLM workflows: prompt templating, Ollama calls, defensive parsing, streaming, and sequential chaining
Documentation