A file-based memory system that implements Tiago Forte's PARA method across three layers: a knowledge graph with atomic YAML facts in Projects/Areas/Resources/Archives folders, daily notes as a raw timeline, and tacit knowledge about user patterns. It's designed for agents that need to remember things across sessions without relying on context windows. The standout feature is the entity system with summary.md for quick loads and items.yaml for detailed facts, plus built-in memory decay and weekly synthesis. Uses qmd for semantic recall instead of grep. If you're building agents that need durable memory and you don't want to roll your own storage layer, this gives you a thoughtful structure out of the box.
npx skills add https://github.com/paperclipai/paperclip --skill para-memory-files