rag-pipelines
Local-first cognitive memory runtime for agent applications.
Smart text chunker for LLM preprocessing (sections → paragraphs → sentences → hard splits).
