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/review-mining.json

View on GitHub
{
  "artifact_kind": "term_prompt_pack",
  "schema_version": 1,
  "artifact_id": "review-mining-jtbd",
  "generated_at": "2026-09-07",
  "locale": "zh-CN",
  "title": "评论挖掘:用 JTBD 把「谁在用、什么痛点」升级成可迭代的产品建议",
  "summary": "你要从一批产品评论里挖产品开发方向。5 个术语里,JTBD 改变分析骨架,ABSA 和 Kano 决定怎么拆和怎么排,VOC 和 Persona 负责收敛成能交接的结果。术语版 prompt 比大白话短一半,AI 会多补三类需求和替代方案。",
  "task": {
    "goal": "从一批产品评论里找出产品开发和迭代方向",
    "known": "已经从第三方工具导出评论数据,知道要看人群、场景、痛点和未满足需求",
    "domain": "消费品产品开发与用户研究",
    "assumptions": [
      "评论数据是中文或英文文本,带评分",
      "目标是迭代现有产品,不是从零定义新品类"
    ]
  },
  "terms": [
    {
      "id": "jtbd",
      "term": "JTBD (Jobs To Be Done)",
      "chinese_name": "待办任务理论",
      "definition": "用户不是在买产品,而是在特定情境下「雇用」产品去完成一项任务;任务包含功能、情感和社交三个维度。",
      "why_for_this_task": "让 AI 从「谁在用、什么场景、什么痛点」的描述式分析,切换到按任务、情境、三类需求和替代方案展开,会补上你没提的维度。",
      "source_status": "verified",
      "sources": [
        {
          "title": "Competing Against Luck: The Story of Innovation and Customer Choice",
          "kind": "book",
          "author": "Clayton M. Christensen 等",
          "year": 2016
        }
      ]
    },
    {
      "id": "kano",
      "term": "Kano Model",
      "chinese_name": "卡诺模型",
      "definition": "把产品属性按用户满意度分成必备、期望、魅力、无差异和反向五类,同一功能的有无对满意度的影响并不对称。",
      "why_for_this_task": "评论里的抱怨和惊喜要分开处理:必备属性缺失会引发差评,魅力属性才拉好评。AI 知道这个词就会给功能分层,而不是把所有痛点平铺。",
      "source_status": "verified",
      "sources": [
        {
          "title": "Attractive Quality and Must-Be Quality",
          "kind": "paper",
          "author": "Noriaki Kano 等",
          "year": 1984
        }
      ]
    },
    {
      "id": "voc",
      "term": "VOC (Voice of the Customer)",
      "chinese_name": "客户之声",
      "definition": "一套把客户原话系统整理成需求层级(一级、二级、三级需求)并按重要度排序的方法。",
      "why_for_this_task": "AI 会把散乱评论收敛成有层级、可排序的需求树,而不是给你一长串并列的关键词。",
      "source_status": "verified",
      "sources": [
        {
          "title": "The Voice of the Customer",
          "kind": "paper",
          "author": "Abbie Griffin, John R. Hauser",
          "year": 1993
        }
      ]
    },
    {
      "id": "persona",
      "term": "Persona",
      "chinese_name": "用户画像",
      "definition": "基于真实用户行为归纳出的典型虚构人物,带有目标、情境和行为特征,用来代表一类用户。",
      "why_for_this_task": "让 AI 把「什么人在用」输出成 2–4 个带目标和情境的画像,而不是「年轻人、宝妈」这类标签。",
      "source_status": "verified",
      "sources": [
        {
          "title": "The Inmates Are Running the Asylum",
          "kind": "book",
          "author": "Alan Cooper",
          "year": 1999
        }
      ]
    },
    {
      "id": "absa",
      "term": "ABSA (Aspect-Based Sentiment Analysis)",
      "chinese_name": "方面级情感分析",
      "definition": "把一条评论拆到具体方面(如电池、噪音、做工),分别判断每个方面的情感极性,而不是给整条评论一个总分。",
      "why_for_this_task": "AI 会按方面统计正负面,你能直接看到哪个部件被骂得最多,这是做产品迭代最直接的输入。",
      "source_status": "verified",
      "sources": [
        {
          "title": "SemEval-2014 Task 4: Aspect Based Sentiment Analysis",
          "kind": "paper",
          "author": "Maria Pontiki 等",
          "year": 2014
        }
      ]
    }
  ],
  "prompt_rewrites": [
    {
      "id": "rw-jtbd",
      "term_ids": [
        "jtbd"
      ],
      "plain_prompt": "附件是我从工具里导出的某品类评论数据。请你帮我深入且系统地分析:这些评论里到底有哪些人群在用?他们在什么场景下使用?使用中有哪些痛点,有哪些需求没有被满足?我的目标是做这个品类的产品开发和迭代,请给我建议。",
      "term_prompt": "附件是我导出的某品类评论数据。请基于 JTBD(Jobs To Be Done)框架分析这些评论,输出产品开发和迭代建议,每条建议附上评论原文作为证据。",
      "added_dimensions": [
        "用户「雇用」产品去完成什么任务,而不只是用哪个功能",
        "功能、情感、社交三类需求分开列",
        "现有替代方案是什么,产品会不会被替代",
        "任务被触发的情境,和使用场景分开看"
      ]
    },
    {
      "id": "rw-absa-kano",
      "term_ids": [
        "absa",
        "kano"
      ],
      "plain_prompt": "请把这些评论里大家抱怨最多的地方和最满意的地方分别列出来,最好能告诉我每个方面有多少人在抱怨、多少人满意,哪些问题最严重、优先改哪个,哪些优点可以继续放大,理由也写一下。",
      "term_prompt": "请对这些评论做 ABSA(方面级情感分析)并统计各方面正负面比例,再按 Kano 模型归类为必备、期望或魅力属性,给出迭代优先级。",
      "added_dimensions": [
        "评论被拆到具体部件或方面,而不是整条打分",
        "每个方面有正负面比例,可以排序",
        "功能按 Kano 分层,必备属性缺失优先修,魅力属性再放大"
      ]
    },
    {
      "id": "rw-voc-persona",
      "term_ids": [
        "voc",
        "persona"
      ],
      "plain_prompt": "帮我看看这些评论都是什么样的人写的,他们各自最在意什么,最好能整理成几类人,每类人的需求分别是什么,需求之间哪个更重要也帮我排一下,分层列出来。",
      "term_prompt": "请按 VOC 方法把评论整理成三级需求树并排序,再归纳 2–4 个 Persona,标出各自最看重的需求分支。",
      "added_dimensions": [
        "需求有层级和排序,不是并列关键词",
        "画像带目标和情境,不是人口标签",
        "每个画像对应到需求树的具体分支"
      ]
    }
  ],
  "learning_path": [
    {
      "term_id": "jtbd",
      "minutes": 10,
      "why_this_order": "它改变整个分析的骨架,先学它,其余四个词都是在它下面补细节。"
    },
    {
      "term_id": "absa",
      "minutes": 6,
      "why_this_order": "决定评论怎么被拆开,是数据处理的第一步。"
    },
    {
      "term_id": "kano",
      "minutes": 6,
      "why_this_order": "拆开之后要分层排优先级,Kano 就是那把尺子。"
    },
    {
      "term_id": "voc",
      "minutes": 5,
      "why_this_order": "把结果收敛成需求树,方便交给产品和团队。"
    },
    {
      "term_id": "persona",
      "minutes": 3,
      "why_this_order": "最后给需求安上人,你已经会用了,只需要知道标准写法。"
    }
  ],
  "unverified": [
    {
      "term": "痛点分级",
      "reason": "多篇文章各有一套分法,没有找到公认出处,本次不列为正式术语。"
    }
  ],
  "evidence_summary": {
    "verified": 5,
    "source_defined": 0,
    "private": 0,
    "unverified": 1,
    "online_verification": true,
    "notes": [
      "五个术语都在 2026-09 联网核验过原始出处。"
    ]
  }
}
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