Forgetful Semantic Memory

Forgetful is a specialized skill that empowers Claude with a persistent knowledge base using atomic note principles. It provides a rigorous framework for deciding when to record information versus when to query existing records, ensuring that architectural decisions, technical patterns, and user preferences are preserved across sessions. By utilizing importance scoring, semantic clustering, and project discovery based on Git metadata, it helps maintain a manageable and highly relevant knowledge graph that prevents duplicate entries and surface-level noise.

Key Features

Use Cases