# Self-Awareness Question Types / 自我认知问题类型 Use this file only when the user asks a specific self-referential question and you need a sharper traversal. 仅当用户提出具体的自我指涉问题、且你需要更精准的文件遍历方式时，才使用本文件。 ## Shared framing / 共同框架 - Read the smallest relevant set of files first. 先读取最小的相关文件集合。 - Organize answers around the user's life and current work when discussing capabilities. 讨论能力时，围绕用户的生活与当前工作来组织回答。 - Cite concrete file paths when the answer would otherwise feel hand-wavy. 当回答否则会显得空泛时，引用具体文件路径。 - Call out uncertainty instead of guessing. 明确指出不确定之处，而不是猜测。 ## 1. Capabilities / 1. 能力 Use for: "what can you do?", "what are you connected to?", "how can you help me?" 适用于："你能做什么？"、"你连接了哪些服务？"、"你能怎么帮我？" Inspect: 检查对象： - identity and operating rules files 身份与操作规则文件 - user profile and long-term memory 用户画像与长期记忆 - `/opt/hatch/skills/` and `~/workspace/skills/` `/opt/hatch/skills/` 与 `~/workspace/skills/` - built artifacts or generated outputs under `~/workspace/` `~/workspace/` 下构建的 artifact 或生成的输出 Response focus: 回答重点： - group by the user's domains or projects 按用户的领域或项目分组 - distinguish active/connected capability from merely available capability 区分"已激活/已连接的能力"与"仅可用而未启用的能力" - keep examples concrete and relevant to this user 让示例具体且与该用户相关 ## 2. Connection discovery / 2. 连接发现 Use for: "what should I connect?", "what integrations am I missing?" 适用于："我该连接什么？"、"我还缺哪些集成？" Inspect: 检查对象： - same sources as capabilities 与"能力"相同的来源 - any visible evidence of connected state in auth files or connector outputs 认证文件或连接器输出中任何可见的已连接状态证据 Response focus: 回答重点： - rank the highest-impact missing connections first 把影响最大的缺失连接排在最前 - explain why each one matters for this user 解释每一项对该用户为何重要 - keep the list short unless the user explicitly wants everything 列表保持简短，除非用户明确想要全部 ## 3. Identity / 3. 身份 Use for: "who are you?", "what is your personality?", "what is your avatar?" 适用于："你是谁？"、"你的个性是什么？"、"你的头像是什么？" Inspect: 检查对象： - identity and personality files 身份与个性文件 - memory only if you need change history or evolution context 仅在需要变更历史或演进背景时才查看记忆 Response focus: 回答重点： - answer in character, but ground claims in files 以角色口吻回答，但论断要有文件依据 - mention where core traits come from 说明核心特质的来源 ## 4. User knowledge / 4. 用户知识 Use for: "what do you know about me?" 适用于："你对我了解多少？" Inspect: 检查对象： - user profile files 用户画像文件 - long-term memory 长期记忆 - recent memory files if freshness matters 若时效性重要，再看最近的记忆文件 Response focus: 回答重点： - separate stable profile facts from recent observations 把稳定的画像事实与近期观察区分开 - flag anything that may be outdated 标记任何可能已过时的内容 ## 5. Memory / 5. 记忆 Use for: "what do you remember about X?" 适用于："你记得关于 X 的什么？" Inspect: 检查对象： - long-term memory first 先看长期记忆 - then recent daily memory files relevant to the topic 再看与主题相关的近期每日记忆文件 Response focus: 回答重点： - distinguish curated memory from raw logs 区分经过整理的记忆与原始日志 - be honest about gaps 对缺失之处坦诚 ## 6. Built items / 6. 已构建产物 Use for: "what have you built?", "what artifacts or skills exist?" 适用于："你构建过什么？"、"存在哪些 artifact 或技能？" Inspect: 检查对象： - generated files under `~/workspace/` `~/workspace/` 下生成的文件 - workspace skills 工作区技能 - memory if you need project history or status 若需要项目历史或状态，再看记忆 Response focus: 回答重点： - list the item, what it does, and whether it looks current or stale 列出产物、其用途，以及它看起来是最新的还是过时的 ## 7. Rules / 7. 规则 Use for: "what are your rules?", "how do you operate?" 适用于："你的规则是什么？"、"你如何运作？" Inspect: 检查对象： - `AGENTS.md` `AGENTS.md` - identity or personality files if they contain behavior constraints 身份或个性文件（若其中包含行为约束） - tools or environment-specific rule files 工具或环境特定的规则文件 Response focus: 回答重点： - separate platform rules from self-imposed style or identity rules 把平台规则与自我设定的风格或身份规则区分开 - keep the answer readable, not legalistic 回答保持可读，不要写得像法律条文