the-space-memory

Use this agent when the user needs thorough knowledge base research that requires multiple search queries, reading full source documents, and synthesizing findings. Trigger when the user says things like "investigate", "research thoroughly", "look into everything about", "find all references to", "deep dive into", or when a simple search isn't sufficient to answer the question. <example> Context: User wants comprehensive information about a topic across multiple documents. user: "前に調べたLoRaの山間部活用について詳しくまとめて" assistant: "I'll use the deep-research agent to thoroughly investigate LoRa usage in mountainous areas across the knowledge base." <commentary> The user wants a thorough investigation that requires multiple queries and reading full documents. </commentary> </example> <example> Context: User wants to understand the history and evolution of a decision. user: "ナレッジ検索の設計判断の経緯を全部洗い出して" assistant: "I'll use the deep-research agent to trace the full history of knowledge search design decisions." <commentary> Tracing history requires multiple time-based queries and cross-referencing documents. </commentary> </example> <example> Context: User asks a question that needs context from multiple sources. user: "射撃とハンドロードについて調べた内容を整理して" assistant: "I'll use the deep-research agent to gather and organize information about shooting and handloading." <commentary> Multiple topics need separate queries and consolidated results. </commentary> </example>

Run health checks on The Space Memory daemon, embedder, and database.

Search the cross-workspace knowledge base using hybrid FTS5 + vector search. Use when: user asks about past research, notes, decisions, or anything that might be in the knowledge base. Examples: "前に調べたLoRaの件どうなってた？", "ナレッジから探して", "〜について調べた記録ある？", "〜あったっけ？", "以前まとめた〜", "Search for anything about X", "What did I write about X?"

3 hooks across 3 events