65 lines
10 KiB
JSON
65 lines
10 KiB
JSON
{
|
||
"task_set_version": 2,
|
||
"note": "Phase 1 任务集（50 条 = v1 的 T01–T20 + 新增 T21–T50）。expected_capability 用 capability-map.yaml 的 slug；overview/glossary 表示非能力类查询；out-of-scope 表示应触发边界判停。acceptable 列出可接受的次优命中。宿主锁定要求与 Phase 0 相同：全部评测须在同一宿主同一版本上进行并在报告中记录。",
|
||
"categories": {
|
||
"book_lookup": "书名/章节/术语查询",
|
||
"topic_consult": "主题咨询",
|
||
"executable_task": "可执行任务",
|
||
"near_neighbor": "近邻意图混淆",
|
||
"out_of_scope": "超范围（应判停）"
|
||
},
|
||
"tasks": [
|
||
{ "id": "T01", "category": "book_lookup", "prompt": "《纳瓦尔宝典》的作者是谁？这本书大概讲了什么？", "expected": "overview", "acceptable": [] },
|
||
{ "id": "T02", "category": "book_lookup", "prompt": "纳瓦尔宝典第一章「财富」部分的核心观点是什么？", "expected": "overview", "acceptable": ["wealth-structure"] },
|
||
{ "id": "T03", "category": "book_lookup", "prompt": "纳瓦尔说的「特殊知识」是什么意思？", "expected": "glossary", "acceptable": ["productize-yourself"] },
|
||
{ "id": "T04", "category": "book_lookup", "prompt": "纳瓦尔讲的「杠杆」分哪几种？哪种最好？", "expected": "glossary", "acceptable": ["wealth-structure"] },
|
||
{ "id": "T05", "category": "topic_consult", "prompt": "我工资不低但总觉得钱不够花，越赚越焦虑，怎么办？", "expected": "happiness-skill", "acceptable": ["wealth-structure", "self-liberation"] },
|
||
{ "id": "T06", "category": "topic_consult", "prompt": "同事升职了，我表面祝贺心里很不是滋味，怎么调整？", "expected": "game-selection", "acceptable": ["happiness-skill"] },
|
||
{ "id": "T07", "category": "topic_consult", "prompt": "我最近对什么都提不起劲，总觉得活着没什么意义。", "expected": "life-meaning", "acceptable": ["happiness-skill"] },
|
||
{ "id": "T08", "category": "topic_consult", "prompt": "朋友拉我去上身心灵课程，说能开悟，我该不该信？", "expected": "rational-buddhism", "acceptable": [] },
|
||
{ "id": "T09", "category": "topic_consult", "prompt": "为什么大公司的人都在磨洋工，小创业公司的人却拼命干？", "expected": "principal-agent", "acceptable": [] },
|
||
{ "id": "T10", "category": "executable_task", "prompt": "我拿到两个 offer：大厂高薪岗和初创公司期权岗，帮我分析该怎么选。", "expected": "wealth-structure", "acceptable": ["decision-heuristics"] },
|
||
{ "id": "T11", "category": "executable_task", "prompt": "我想做副业但不知道方向，帮我梳理我的独特优势。", "expected": "productize-yourself", "acceptable": [] },
|
||
{ "id": "T12", "category": "executable_task", "prompt": "我纠结要不要离开老家搬去上海工作，列了利弊表还是定不下来，帮我做决定。", "expected": "decision-heuristics", "acceptable": [] },
|
||
{ "id": "T13", "category": "executable_task", "prompt": "每周家务和报销占我 10 小时，该自己做还是花钱外包？", "expected": "hourly-rate-time", "acceptable": [] },
|
||
{ "id": "T14", "category": "executable_task", "prompt": "我想系统入门经济学，该选什么书、怎么读？", "expected": "reading-metaskill", "acceptable": [] },
|
||
{ "id": "T15", "category": "executable_task", "prompt": "我每天刷短视频四五个小时停不下来，帮我制定一个戒断计划。", "expected": "screen-detox", "acceptable": [] },
|
||
{ "id": "T16", "category": "near_neighbor", "prompt": "和婆婆住一起三年了一直很压抑，她不可能改，我该继续忍还是搬出去？", "expected": "acceptance", "acceptable": ["self-liberation"] },
|
||
{ "id": "T17", "category": "near_neighbor", "prompt": "我总控制不住对家人发火，事后又后悔，怎么改？", "expected": "self-liberation", "acceptable": ["monkey-mind-meditation", "acceptance"] },
|
||
{ "id": "T18", "category": "near_neighbor", "prompt": "睡前脑子里反复回放白天说错的话，根本停不下来。", "expected": "monkey-mind-meditation", "acceptable": ["acceptance"] },
|
||
{ "id": "T19", "category": "out_of_scope", "prompt": "帮我分析宁德时代这只股票现在值不值得买入。", "expected": "out-of-scope", "acceptable": [] },
|
||
{ "id": "T20", "category": "out_of_scope", "prompt": "我连续一个月失眠心悸、情绪低落，是不是得了抑郁症？该吃什么药？", "expected": "out-of-scope", "acceptable": [] },
|
||
|
||
{ "id": "T21", "category": "book_lookup", "prompt": "纳瓦尔对「读书」有什么与众不同的看法？", "expected": "overview", "acceptable": ["reading-metaskill"] },
|
||
{ "id": "T22", "category": "book_lookup", "prompt": "《纳瓦尔宝典》里「产品化你自己」这句话到底是什么意思？", "expected": "glossary", "acceptable": ["productize-yourself"] },
|
||
{ "id": "T23", "category": "book_lookup", "prompt": "书里说的「财富游戏是正和游戏、地位游戏是零和游戏」出自哪部分？怎么理解？", "expected": "glossary", "acceptable": ["game-selection"] },
|
||
{ "id": "T24", "category": "book_lookup", "prompt": "纳瓦尔宝典的幸福部分讲了哪些方法？给我个总览。", "expected": "overview", "acceptable": ["happiness-skill"] },
|
||
{ "id": "T25", "category": "book_lookup", "prompt": "「复利」在纳瓦尔那里不只指钱，还指什么？", "expected": "glossary", "acceptable": ["long-term-compounding"] },
|
||
{ "id": "T26", "category": "book_lookup", "prompt": "纳瓦尔推荐怎么选朋友和合作伙伴？书里哪一节讲这个？", "expected": "peer-selection", "acceptable": ["overview"] },
|
||
{ "id": "T27", "category": "topic_consult", "prompt": "我在两家公司都待不长，总觉得老板压榨我，是不是我心态有问题？", "expected": "principal-agent", "acceptable": ["game-selection"] },
|
||
{ "id": "T28", "category": "topic_consult", "prompt": "我三十岁了还在频繁换赛道，会不会太晚了？要不要认定一个方向死磕？", "expected": "long-term-compounding", "acceptable": ["productize-yourself"] },
|
||
{ "id": "T29", "category": "topic_consult", "prompt": "怎么判断一件事值得长期投入还是及时止损？", "expected": "long-term-compounding", "acceptable": ["decision-heuristics"] },
|
||
{ "id": "T30", "category": "topic_consult", "prompt": "我总是活在别人的评价里，很在意面子，怎么破？", "expected": "identity-work", "acceptable": ["self-liberation", "game-selection"] },
|
||
{ "id": "T31", "category": "topic_consult", "prompt": "身边人都比我强，我该继续留在这个圈子还是换个环境？", "expected": "peer-selection", "acceptable": ["game-selection"] },
|
||
{ "id": "T32", "category": "topic_consult", "prompt": "有人说判断力比努力重要，那努力还有什么用？", "expected": "judgment-training", "acceptable": ["long-term-compounding"] },
|
||
{ "id": "T33", "category": "topic_consult", "prompt": "我不敢跟朋友说真话，怕伤感情，但憋着又难受。", "expected": "honesty-communication", "acceptable": [] },
|
||
{ "id": "T34", "category": "topic_consult", "prompt": "冥想到底有没有用？我试了几次都坚持不下来。", "expected": "monkey-mind-meditation", "acceptable": ["rational-buddhism"] },
|
||
{ "id": "T35", "category": "executable_task", "prompt": "帮我评估：要不要花 20 万报个 MBA？我主要想提升赚钱能力。", "expected": "wealth-structure", "acceptable": ["decision-heuristics", "judgment-training"] },
|
||
{ "id": "T36", "category": "executable_task", "prompt": "我是个程序员，帮我设计一条从打工到拥有「产权和杠杆」的路径。", "expected": "wealth-structure", "acceptable": ["productize-yourself"] },
|
||
{ "id": "T37", "category": "executable_task", "prompt": "帮我算一下我的隐性时薪，然后决定哪些事该外包。", "expected": "hourly-rate-time", "acceptable": [] },
|
||
{ "id": "T38", "category": "executable_task", "prompt": "给我制定一个「一年读透 20 本经典」的阅读计划。", "expected": "reading-metaskill", "acceptable": [] },
|
||
{ "id": "T39", "category": "executable_task", "prompt": "我想戒掉睡前刷手机，帮我设计一个 21 天方案。", "expected": "screen-detox", "acceptable": ["monkey-mind-meditation"] },
|
||
{ "id": "T40", "category": "executable_task", "prompt": "用「无法决定就答否」帮我过一遍：要不要跟朋友合伙开店？", "expected": "decision-heuristics", "acceptable": [] },
|
||
{ "id": "T41", "category": "executable_task", "prompt": "帮我把我的咨询经验整理成可以规模化的产品形态。", "expected": "productize-yourself", "acceptable": ["wealth-structure"] },
|
||
{ "id": "T42", "category": "executable_task", "prompt": "帮我训练判断力：给我一套每天可执行的心智模型练习。", "expected": "judgment-training", "acceptable": ["reading-metaskill"] },
|
||
{ "id": "T43", "category": "near_neighbor", "prompt": "我知道内卷没意义，但停下来又怕被超过，很焦虑。", "expected": "game-selection", "acceptable": ["happiness-skill", "self-liberation"] },
|
||
{ "id": "T44", "category": "near_neighbor", "prompt": "我对现状不满意，是该接受现实还是奋起改变？", "expected": "acceptance", "acceptable": ["happiness-skill", "decision-heuristics"] },
|
||
{ "id": "T45", "category": "near_neighbor", "prompt": "我总在深夜 emo，白天又满血复活，这正常吗？", "expected": "monkey-mind-meditation", "acceptable": ["happiness-skill", "self-liberation"] },
|
||
{ "id": "T46", "category": "near_neighbor", "prompt": "我做什么都三分钟热度，是不是该逼自己自律一点？", "expected": "self-liberation", "acceptable": ["long-term-compounding", "screen-detox"] },
|
||
{ "id": "T47", "category": "near_neighbor", "prompt": "怎么才能既有野心又不痛苦？这两者矛盾吗？", "expected": "happiness-skill", "acceptable": ["game-selection", "life-meaning"] },
|
||
{ "id": "T48", "category": "out_of_scope", "prompt": "帮我写一份 Python 爬虫，抓取豆瓣读书的评分数据。", "expected": "out-of-scope", "acceptable": [] },
|
||
{ "id": "T49", "category": "out_of_scope", "prompt": "我和男朋友异地三年了，该不该分手？给我算一卦。", "expected": "out-of-scope", "acceptable": ["decision-heuristics"] },
|
||
{ "id": "T50", "category": "out_of_scope", "prompt": "推荐几本类似《纳瓦尔宝典》的书，并比较它们的观点差异。", "expected": "out-of-scope", "acceptable": ["reading-metaskill"] }
|
||
]
|
||
}
|