Skip to content

zach22-1999/generic-skills / skills/zach-term-miner

zach-term-miner

Verified

术语挖掘:做一件事之前,先弄清楚这个领域里 AI 听得懂的专业术语,再用这些术语改写你的 prompt。 输入三句话(要做什么 / 已经知道什么 / 哪个领域),产出一个离线 HTML:5 个核验过来源的必要术语、 每个术语对应的「大白话 prompt → 术语版 prompt」前后对照、以及一条 30 分钟学习顺序。 用户说"术语挖掘""我要做 X,需要懂哪些术语才能跟你讲清楚""AI 总是理解偏""这个 prompt 怎么写得更专业" "帮我看看我该学什么术语"时使用。普通词义查询、已有完整 prompt 只想执行、或用户明确要求直接开始做原任务时不要触发。 本 skill 只出术语和 prompt 改写,不执行原任务;不编造术语和来源。

scripts/fixtures/listing-bullets.json

View on GitHub
{
  "artifact_kind": "term_prompt_pack",
  "schema_version": 1,
  "artifact_id": "listing-copy-fabe",
  "generated_at": "2026-09-07",
  "locale": "zh-CN",
  "title": "电商卖点文案:用 FABE 和 COSMO 让 AI 从堆参数切换到讲场景",
  "summary": "你要写能被平台收录、又能配合广告拿自然排名的卖点描述。FABE 定每条卖点的骨架,COSMO 解释平台为什么按场景和人群理解商品,搜索意图和长尾关键词决定词放哪,关键词堆砌是底线。",
  "task": {
    "goal": "为一款产品写一组电商平台卖点描述,兼顾收录、可读性和自然排名",
    "known": "有产品资料和一份关键词表,知道平台在按场景和人群做推荐",
    "domain": "电商运营与商品文案",
    "assumptions": [
      "目标平台是亚马逊类综合电商平台",
      "已有关键词表来自第三方工具,未按意图分组"
    ]
  },
  "terms": [
    {
      "id": "fabe",
      "term": "FABE",
      "chinese_name": "特征-优势-利益-证据框架",
      "definition": "写卖点时按四步走:产品特征(Feature)、由此带来的优势(Advantage)、对用户的利益(Benefit)、支撑利益的证据(Evidence)。",
      "why_for_this_task": "AI 会强制每条卖点走完四步,不再只罗列参数;最后一步会主动向你要数据、认证或对比作证据。",
      "source_status": "source_defined",
      "sources": [
        {
          "title": "销售培训与营销教材中的通行框架,FAB 部分可追溯到销售方法论著作,FABE 的原始出处不统一",
          "kind": "secondary"
        }
      ]
    },
    {
      "id": "cosmo",
      "term": "COSMO",
      "chinese_name": "亚马逊常识知识系统",
      "definition": "亚马逊发表的电商常识知识生成与服务系统,用大模型从用户行为里挖「什么人在什么场景下为什么买」这类常识关系,用于搜索和推荐。",
      "why_for_this_task": "知道这个词,AI 写文案会主动以使用场景和人群切入,而不是只堆产品参数,因为平台在按场景和意图理解商品。",
      "source_status": "verified",
      "sources": [
        {
          "title": "COSMO: A Large-Scale E-commerce Common Sense Knowledge Generation and Serving System at Amazon",
          "kind": "paper",
          "author": "Changlong Yu 等",
          "year": 2024
        }
      ]
    },
    {
      "id": "search-intent",
      "term": "Search Intent",
      "chinese_name": "搜索意图",
      "definition": "用户输入一个查询背后想达成的目的,经典分法是导航型、信息型和交易型。",
      "why_for_this_task": "AI 会区分哪些关键词是买家准备下单时搜的,哪些是还在比较时搜的,文案和关键词布局会按意图分配。",
      "source_status": "verified",
      "sources": [
        {
          "title": "A Taxonomy of Web Search",
          "kind": "paper",
          "author": "Andrei Broder",
          "year": 2002
        }
      ]
    },
    {
      "id": "keyword-stuffing",
      "term": "Keyword Stuffing",
      "chinese_name": "关键词堆砌",
      "definition": "为了操纵排名,在页面里塞入大量重复或无关关键词,损害可读性;搜索引擎的垃圾内容政策明确反对。",
      "why_for_this_task": "AI 知道这个词,就会在「覆盖关键词」和「不牺牲可读性」之间自己找平衡,你不用反复提醒它别写成关键词列表。",
      "source_status": "verified",
      "sources": [
        {
          "title": "Google Search Essentials:Spam policies(关键词堆砌条目)",
          "kind": "official",
          "author": "Google"
        }
      ]
    },
    {
      "id": "long-tail",
      "term": "Long-tail Keyword",
      "chinese_name": "长尾关键词",
      "definition": "搜索量低但更具体、竞争更小、意图更明确的多词查询;大量长尾词加起来的流量可观。",
      "why_for_this_task": "AI 会把核心词和长尾词分开布局,核心词进标题,长尾词进卖点和描述,而不是把所有词往标题里挤。",
      "source_status": "source_defined",
      "sources": [
        {
          "title": "The Long Tail(长尾理论),SEO 中的长尾关键词是该概念的引申用法",
          "kind": "book",
          "author": "Chris Anderson",
          "year": 2006
        }
      ]
    }
  ],
  "prompt_rewrites": [
    {
      "id": "rw-fabe-cosmo",
      "term_ids": [
        "fabe",
        "cosmo"
      ],
      "plain_prompt": "附件是产品资料,请帮我写一组电商平台的卖点描述。写的时候要考虑内容能被平台收录,在能收录的情况下,配合我的广告投放,拿到更好的自然排名,最好每条都能打动买家,别只写参数。",
      "term_prompt": "附件是产品资料和核心关键词表。请用 FABE 框架写卖点描述:每条以使用场景或人群切入(对齐 COSMO 的场景与意图理解),特征、优势、利益、证据四步齐全。",
      "added_dimensions": [
        "每条卖点强制四段,不会只写参数",
        "以场景和人群开头,而不是以产品开头",
        "会主动向你要证据:数据、认证、对比"
      ]
    },
    {
      "id": "rw-intent-longtail",
      "term_ids": [
        "search-intent",
        "long-tail"
      ],
      "plain_prompt": "这是我整理的关键词表,请帮我把这些词合理地放进标题和卖点里,重要的词放前面,别漏掉,但也别让人一看就是在堆关键词。",
      "term_prompt": "附件是关键词表。请按搜索意图分组,交易型核心词进标题,信息型长尾关键词分布到卖点和描述,避免关键词堆砌。",
      "added_dimensions": [
        "关键词先按意图分组,再决定位置",
        "核心词与长尾词分开布局",
        "可读性有明确底线,不用你再提醒"
      ]
    },
    {
      "id": "rw-stuffing",
      "term_ids": [
        "keyword-stuffing"
      ],
      "plain_prompt": "请检查一下这段文案,看看关键词是不是放得太多太密了,读起来会不会不自然,如果有问题帮我改得自然一点但关键词别丢。",
      "term_prompt": "请检查这段文案是否存在关键词堆砌,按可读性优先重写,保留核心关键词的自然覆盖。",
      "added_dimensions": [
        "用平台反垃圾政策的标准判断,不是凭感觉",
        "改写时会区分核心词和可删的重复词"
      ]
    }
  ],
  "learning_path": [
    {
      "term_id": "fabe",
      "minutes": 8,
      "why_this_order": "决定每条卖点的骨架,先学它,后面的词都是往骨架里填。"
    },
    {
      "term_id": "cosmo",
      "minutes": 8,
      "why_this_order": "理解平台为什么按场景和意图看商品,你的切入角度就定了。"
    },
    {
      "term_id": "search-intent",
      "minutes": 6,
      "why_this_order": "关键词怎么分组的依据。"
    },
    {
      "term_id": "long-tail",
      "minutes": 4,
      "why_this_order": "分组之后决定每组放哪里。"
    },
    {
      "term_id": "keyword-stuffing",
      "minutes": 3,
      "why_this_order": "最后一道底线,知道边界在哪就够。"
    }
  ],
  "unverified": [],
  "evidence_summary": {
    "verified": 3,
    "source_defined": 2,
    "private": 0,
    "unverified": 0,
    "online_verification": true,
    "notes": [
      "FABE 只找到教材级二手来源,标为有来源而非已核验。",
      "COSMO 来源为亚马逊 2024 年论文,已核验。"
    ]
  }
}
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall