[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M72YGW06SYRH0GF86007BSA2":3,"skill-same:01M72YGW06SYRH0GF86007BSA2":58},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M72YGW06SYRH0GF86007BSA2","gbrain-tcm-enrich",null,"使用 gbrain 官方 enrich 流程，從中醫經典文本（如倪海厦《金匱要略》視頻同步文稿）建立知識圖譜。\n適用於：處理中醫文獻、建立倪師臨床經驗知識庫、對中藥\u002F方劑\u002F病理進行實體蒸餾。\n觸發時機：用户要求「處理中醫文本」、「建立倪師知識圖譜」、「 enrich 這份中醫文獻」。","Morris-UTRUST\u002FGBrain-knowledge","Morris-UTRUST","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F13375437?v=4","01M72YEA3A0GZ2EQY0ZXRVEB7E","陳冠榮","01M72YEA3VYSH6HF42JCT3T7HA",0,"MIT","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Fskills\u002Fgbrain-tcm-enrich\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: gbrain-tcm-enrich\ndescription: |\n  使用 gbrain 官方 enrich 流程，從中醫經典文本（如倪海厦《金匱要略》視頻同步文稿）建立知識圖譜。\n  適用於：處理中醫文獻、建立倪師臨床經驗知識庫、對中藥\u002F方劑\u002F病理進行實體蒸餾。\n  觸發時機：用户要求「處理中醫文本」、「建立倪師知識圖譜」、「 enrich 這份中醫文獻」。\n---\n\n# Gbrain TCM Enrich — 中醫文本知識蒸餾工作流\n\n## 核心理念\n\n**不是處理文本，是從文本中發現地圖。**\n\n文本只是證據來源，不是目標產物。真正的目標是：每提到一個中醫實體（單味藥、方劑、病理），就建立一個有 citations、有 back-link、有 timeline 的 brain page。\n\n## 工作流程（4步）\n\n### Step 1：建立核心 anchor\n\n第一個動作：確認 `people\u002F倪海厦` 存在。\n\n```bash\ncd ~\u002Fgbrain && ~\u002F.bun\u002Fbin\u002Fbun run src\u002Fcli.ts list 2>&1 | grep -v ai.gateway | grep \"people\u002F倪海厦\"\n```\n\n若不存在，用 `put people\u002F倪海厦` 建立。這個 page 是所有其他實體的 back-link 終點。\n\n### Step 2：分析文本結構（不是 enrich，先了解範圍）\n\n在開始 enrich 前，先對文本做結構成像：\n\n1. **確認文本類型**：是「視頻同步文稿」還是「原始文獻」？\n2. **確認目錄結構**（適用於《金匱要略》這類有篇目分類的文本）：\n   ```bash\n   grep -n \"篇\\|第.*章\\|证治\\|病脉\" \u002Ftmp\u002Fgbrain-backup\u002F倪海厦\u002F*.md | head -30\n   ```\n3. **確認實體密度**：哪些章節提到單味藥？哪些提到方劑？哪些提到病理？\n\n這個步驟用 `grep` 即可，不需要建立任何 brain page。\n\n### Step 3：系統性 Enrich（官方 7 步法自訂版）\n\n每次 enrich 一個實體，順序如下：\n\n**優先順序（Tier 1 先）：**\n1. `people\u002F倪海厦`（核心，必須存在）\n2. 單味藥（herbs\u002F）：臨床最常用的 30 味\n3. 方劑（formulas\u002F）：金匱要略中出現的經方\n4. 中醫病理（conditions\u002F）：瘀血、寒疝、狐惑、痙病等\n\n**Enrich 流程（每次一個實體）：**\n\n```\n1. Identify：從文本中找出該實體的臨床判斷句\n2. Check brain：gbrain list 確認是否已存在\n3. Extract signal：\n   - 臨床判斷（倪師怎麼說）\n   - 藥物组合（對什麼症用什麼藥）\n   - 劑量與煎煮法\n4. Write page：使用 gbrain put 寫入\n5. Create bidirectional link：\n   - 新頁 → people\u002F倪海厦（back-link）\n   - people\u002F倪海厦 → 新頁（update the person page）\n6. Add timeline entry：標注「倪師在金匱要略第X篇談到此實體」\n```\n\n**Citation 格式（必須）：**\n```\n[Source: 金匱要略-0X.md:行號]\n```\n行號從 `grep -n` 的輸出中取得。\n\n### Step 4：Maintain Health（每完成 10 個實體做一次）\n\n```bash\ncd ~\u002Fgbrain && ~\u002F.bun\u002Fbin\u002Fbun run src\u002Fcli.ts extract links --source db\ncd ~\u002Fgbrain && ~\u002F.bun\u002Fbin\u002Fbun run src\u002Fcli.ts health\n```\n\n確認：\n- orphan pages ≤ 5\n- dead links = 0\n- link coverage ≥ 95%\n\n---\n\n## 實體命名規則\n\n| 類型 | 目錄 | 範例 |\n|------|------|------|\n| 中醫人物 | people\u002F | people\u002F倪海厦 |\n| 單味藥 | herbs\u002F | herbs\u002F附子、herbs\u002F桂枝 |\n| 方劑 | formulas\u002F | formulas\u002F四逆湯、formulas\u002F桂枝湯 |\n| 中醫病理 | conditions\u002F | conditions\u002F瘀血、conditions\u002F寒疝 |\n| 概念\u002F入口頁 | concepts\u002F | concepts\u002F倪師臨床地圖 |\n\n**禁止：**\n- 不要用西醫病名作為病理 page 名（如「肝癌」、「肺癌」）\n- 同一實體不可同時存在於兩個目錄（如 `pathology\u002F` 和 `conditions\u002F` 不能都有「寒疝」）\n\n---\n\n## 紀錄規則（Filing + Quality Convention）\n\n1. **Primary subject 決定目錄**：一篇關於附子的臨床筆記 → `herbs\u002F附子`，不是 `sources\u002F` 或 `notes\u002F`\n2. **Inline citation 必須有**：每個臨床判斷事實後面加 `[Source: 金匱要略-0X.md:行號]`\n3. **Back-link 雙向**：每次建立實體 page → link 回 `people\u002F倪海厦`；同步 update `people\u002F倪海厦` 的 linked entities 列表\n4. **Timeline 只能追加**：新資訊加在 timeline 底部，不修改既有的 compiled truth\n\n---\n\n## 參考文獻（使用前先讀）\n\n- [gbrain-filing-rules.md](references\u002Fgbrain-filing-rules.md) — 目錄分類規則，必須遵守\n- [gbrain-quality.md](references\u002Fgbrain-quality.md) — Citation 格式與 back-link Iron Law\n- [gbrain-brain-first.md](references\u002Fgbrain-brain-first.md) — 查詢順序：search → query → get → 外部 API\n\n---\n\n## 工作終止條件\n\n滿足以下任一條件即停止，不需要建立所有可能的實體：\n\n1. **覆蓋充分**：主要章節（痙病、狐惑、瘀血、寒疝、陰陽毒、百合）都已建立實體\n2. **新發現減少**：連續 3 個實體都只來自同一篇，無新篇目\n3. **Brain 健康**：orphan ≤ 5，dead links = 0，agent 即可宣告完成\n\n---\n\n## 常見錯誤（不要这样做）\n\n| 錯誤做法 | 正確做法 |\n|---------|---------|\n| 直接對文本 grep 建立 page | 先 `gbrain list` 確認不存在再建立 |\n| 所有 herbs 一次性建立 | 每次一個，完成並 link 後再下一個 |\n| 混用 conditions\u002F 和 pathology\u002F | 統一用 conditions\u002F |\n| 不加 citation | 每個臨床判斷必須有 `[Source: ...]` |\n| 不連回 people\u002F倪海厦 | 每個新實體 page 必須 link 到 anchor |","# Gbrain TCM Enrich — 中醫文本知識蒸餾工作流\n\n## 核心理念\n\n**不是處理文本，是從文本中發現地圖。**\n\n文本只是證據來源，不是目標產物。真正的目標是：每提到一個中醫實體（單味藥、方劑、病理），就建立一個有 citations、有 back-link、有 timeline 的 brain page。\n\n## 工作流程（4步）\n\n### Step 1：建立核心 anchor\n\n第一個動作：確認 `people\u002F倪海厦` 存在。\n\n```bash\ncd ~\u002Fgbrain && ~\u002F.bun\u002Fbin\u002Fbun run src\u002Fcli.ts list 2>&1 | grep -v ai.gateway | grep \"people\u002F倪海厦\"\n```\n\n若不存在，用 `put people\u002F倪海厦` 建立。這個 page 是所有其他實體的 back-link 終點。\n\n### Step 2：分析文本結構（不是 enrich，先了解範圍）\n\n在開始 enrich 前，先對文本做結構成像：\n\n1. **確認文本類型**：是「視頻同步文稿」還是「原始文獻」？\n2. **確認目錄結構**（適用於《金匱要略》這類有篇目分類的文本）：\n   ```bash\n   grep -n \"篇\\|第.*章\\|证治\\|病脉\" \u002Ftmp\u002Fgbrain-backup\u002F倪海厦\u002F*.md | head -30\n   ```\n3. **確認實體密度**：哪些章節提到單味藥？哪些提到方劑？哪些提到病理？\n\n這個步驟用 `grep` 即可，不需要建立任何 brain page。\n\n### Step 3：系統性 Enrich（官方 7 步法自訂版）\n\n每次 enrich 一個實體，順序如下：\n\n**優先順序（Tier 1 先）：**\n1. `people\u002F倪海厦`（核心，必須存在）\n2. 單味藥（herbs\u002F）：臨床最常用的 30 味\n3. 方劑（formulas\u002F）：金匱要略中出現的經方\n4. 中醫病理（conditions\u002F）：瘀血、寒疝、狐惑、痙病等\n\n**Enrich 流程（每次一個實體）：**\n\n```\n1. Identify：從文本中找出該實體的臨床判斷句\n2. Check brain：gbrain list 確認是否已存在\n3. Extract signal：\n   - 臨床判斷（倪師怎麼說）\n   - 藥物组合（對什麼症用什麼藥）\n   - 劑量與煎煮法\n4. Write page：使用 gbrain put 寫入\n5. Create bidirectional link：\n   - 新頁 → people\u002F倪海厦（back-link）\n   - people\u002F倪海厦 → 新頁（update the person page）\n6. Add timeline entry：標注「倪師在金匱要略第X篇談到此實體」\n```\n\n**Citation 格式（必須）：**\n```\n[Source: 金匱要略-0X.md:行號]\n```\n行號從 `grep -n` 的輸出中取得。\n\n### Step 4：Maintain Health（每完成 10 個實體做一次）\n\n```bash\ncd ~\u002Fgbrain && ~\u002F.bun\u002Fbin\u002Fbun run src\u002Fcli.ts extract links --source db\ncd ~\u002Fgbrain && ~\u002F.bun\u002Fbin\u002Fbun run src\u002Fcli.ts health\n```\n\n確認：\n- orphan pages ≤ 5\n- dead links = 0\n- link coverage ≥ 95%\n\n---\n\n## 實體命名規則\n\n| 類型 | 目錄 | 範例 |\n|------|------|------|\n| 中醫人物 | people\u002F | people\u002F倪海厦 |\n| 單味藥 | herbs\u002F | herbs\u002F附子、herbs\u002F桂枝 |\n| 方劑 | formulas\u002F | formulas\u002F四逆湯、formulas\u002F桂枝湯 |\n| 中醫病理 | conditions\u002F | conditions\u002F瘀血、conditions\u002F寒疝 |\n| 概念\u002F入口頁 | concepts\u002F | concepts\u002F倪師臨床地圖 |\n\n**禁止：**\n- 不要用西醫病名作為病理 page 名（如「肝癌」、「肺癌」）\n- 同一實體不可同時存在於兩個目錄（如 `pathology\u002F` 和 `conditions\u002F` 不能都有「寒疝」）\n\n---\n\n## 紀錄規則（Filing + Quality Convention）\n\n1. **Primary subject 決定目錄**：一篇關於附子的臨床筆記 → `herbs\u002F附子`，不是 `sources\u002F` 或 `notes\u002F`\n2. **Inline citation 必須有**：每個臨床判斷事實後面加 `[Source: 金匱要略-0X.md:行號]`\n3. **Back-link 雙向**：每次建立實體 page → link 回 `people\u002F倪海厦`；同步 update `people\u002F倪海厦` 的 linked entities 列表\n4. **Timeline 只能追加**：新資訊加在 timeline 底部，不修改既有的 compiled truth\n\n---\n\n## 參考文獻（使用前先讀）\n\n- [gbrain-filing-rules.md](references\u002Fgbrain-filing-rules.md) — 目錄分類規則，必須遵守\n- [gbrain-quality.md](references\u002Fgbrain-quality.md) — Citation 格式與 back-link Iron Law\n- [gbrain-brain-first.md](references\u002Fgbrain-brain-first.md) — 查詢順序：search → query → get → 外部 API\n\n---\n\n## 工作終止條件\n\n滿足以下任一條件即停止，不需要建立所有可能的實體：\n\n1. **覆蓋充分**：主要章節（痙病、狐惑、瘀血、寒疝、陰陽毒、百合）都已建立實體\n2. **新發現減少**：連續 3 個實體都只來自同一篇，無新篇目\n3. **Brain 健康**：orphan ≤ 5，dead links = 0，agent 即可宣告完成\n\n---\n\n## 常見錯誤（不要这样做）\n\n| 錯誤做法 | 正確做法 |\n|---------|---------|\n| 直接對文本 grep 建立 page | 先 `gbrain list` 確認不存在再建立 |\n| 所有 herbs 一次性建立 | 每次一個，完成並 link 後再下一個 |\n| 混用 conditions\u002F 和 pathology\u002F | 統一用 conditions\u002F |\n| 不加 citation | 每個臨床判斷必須有 `[Source: ...]` |\n| 不連回 people\u002F倪海厦 | 每個新實體 page 必須 link 到 anchor |",{"name":5,"description":7},"main","d5a406455378d589b645b3744f2ceae49d0b6e23","skills\u002Fgbrain-tcm-enrich","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST","2026-05-17 23:07:15",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add Morris-UTRUST\u002FGBrain-knowledge --skill gbrain-tcm-enrich",{"tool":36,"command":37},"bunx","bunx skills add Morris-UTRUST\u002FGBrain-knowledge --skill gbrain-tcm-enrich",{"tool":39,"command":40},"pnpm","pnpm dlx skills add Morris-UTRUST\u002FGBrain-knowledge --skill gbrain-tcm-enrich","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Fskills\u002Fgbrain-tcm-enrich\u002FSKILL.md",true,[44,50,54],{"path":45,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":48,"sourceHtmlUrl":49},"references\u002Fgbrain-brain-first.md","REFERENCE","参考文档",2129,"https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Fskills\u002Fgbrain-tcm-enrich\u002Freferences\u002Fgbrain-brain-first.md",{"path":51,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":52,"sourceHtmlUrl":53},"references\u002Fgbrain-filing-rules.md",2649,"https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Fskills\u002Fgbrain-tcm-enrich\u002Freferences\u002Fgbrain-filing-rules.md",{"path":55,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Fgbrain-quality.md",2736,"https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Fskills\u002Fgbrain-tcm-enrich\u002Freferences\u002Fgbrain-quality.md",[59,66,73,80,87,94],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":19},"01M72YKMKH4254SNHQZ5PNHDT9","signal-detector","Always-on ambient signal capture for the OpenClaw reference fixture.","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fopenclaw-reference-minimal\u002Fskills\u002Fsignal-detector\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":19},"01M72YKJXCNATARDEE5E1GKGV4","query","Look up brain pages in the OpenClaw reference fixture.","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fopenclaw-reference-minimal\u002Fskills\u002Fquery\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":19},"01M72YKH458S74JP3C0TVM9VW8","context-now","ALWAYS-ON time-sensitivity discipline for the OpenClaw reference fixture.","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fopenclaw-reference-minimal\u002Fskills\u002Fcontext-now\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":19},"01M72YKFAB0TMQ96YB1K1ZSADE","brain-ops","Core read\u002Fwrite cycle for the OpenClaw reference fixture.","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fopenclaw-reference-minimal\u002Fskills\u002Fbrain-ops\u002FSKILL.md",[],[],{"skillNo":88,"name":89,"nameZh":6,"description":90,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":19},"01M72YKC9DHWBE0TPH75KGN55Q","webhook-transforms","Generic framework for converting external events (SMS, meetings, social mentions)\ninto brain-ingestible signals. Define a transform function, register a webhook URL,\nand incoming events get processed through the brain pipeline.","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Fskills\u002Fwebhook-transforms\u002FSKILL.md",[],[],{"skillNo":95,"name":96,"nameZh":6,"description":97,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":19},"01M72YK9J860QDTNRSVKBDA9M6","voice-note-ingest","Ingest a voice note with exact-phrasing preservation (never paraphrased). Routes content to originals\u002F, concepts\u002F, people\u002F, companies\u002F, ideas\u002F, personal\u002F, or voice-notes\u002F based on a decision tree. The user's exact words are the signal.","https:\u002F\u002Fgithub.com\u002FMorris-UTRUST\u002FGBrain-knowledge\u002Fblob\u002Fmain\u002Fskills\u002Fvoice-note-ingest\u002FSKILL.md",[],[]]