概要
Memory Search provides a persistent memory layer for Claude, allowing it to bridge the gap between separate coding sessions. By querying a dedicated memory database, Claude can answer questions about previous bug fixes, implementation details, and architectural decisions that would otherwise be lost when a session ends. The skill employs a token-efficient workflow that first retrieves a lightweight index of results, allowing the user or AI to selectively fetch deep context only when necessary, ensuring historical consistency across long-term projects.
