agentsclimarketplace

Profile intake

Skill tywinlu1988/FolioPulse/.claude/skills/profile-intake

FolioPulse 客户画像摄入技能。当客户经理描述客户情况、提及投资需求、 或询问"帮客户看看""给客户推荐"等意图时触发。 通过预检 + 4 问路由协议采集客户画像,生成结构化 Profile Sheet (YAML)。From its SKILL.md

Install
npx -y skills add tywinlu1988/FolioPulse --skill profile-intake

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

  • 29 days oldThe repository was created 29 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.5 KB, ~1.3k tokens by cl100k_base, as published. Nobody here has run it

用途

将客户经理的自然语言描述转换为结构化的客户画像表(Profile Sheet YAML), 作为后续推荐引擎的输入。

调用协议

§0 预检:信息覆盖度评估

收到客户经理输入后,先不提问。 扫描已有信息,对照画像表必填字段提取:

字段提取关键词/模式
风险等级R1/R2/R3/R4/R5、保守/稳健/平衡/进取/激进
投资金额数字 + 万/元/亿
投资期限短期/中期/长期、<1年/1-3年/>3年、半年/几个月
投资目标增值/保本/养老/教育/流动性/投机
客户姓名"XX客户""XX先生/女士""客户叫XX"
客户经理"我是XX""我叫XX"(默认为当前用户)
投资者类型普通投资者/合格投资者、资产/收入(若提及 ≥500万或 ≥50万年收入)
特殊约束"不投XX""偏好XX""禁投XX""只要XX""不要XX"
流动性需求"随时赎回""可锁定XX月/年""不急用""随时要用"

规则:

  • 提取到的字段,在后续提问中跳过
  • 未提取到的必填字段,按 §1 的四问顺序逐一提问
  • 如果 RM 提供的信息不足以提取任何必填字段,从 Q1 开始
  • 如果所有必填字段均已覆盖,跳过提问,直接生成画像表

§1 逐问采集

仅提问预检中标记为缺失的问题。一次一问。

问序问题选项
Q1客户的投资目标是什么?资产增值 / 稳健收益 / 养老规划 / 子女教育 / 流动性管理 / 其他
Q2风险承受等级?R1(保守)/ R2(稳健)/ R3(平衡)/ R4(进取)/ R5(激进)
Q3计划投资金额与期限?金额(万元)+ 短期(<1年) / 中期(1-3年) / 长期(>3年)
Q4是否有特殊约束?行业偏好 / 禁投行业 / 流动性要求 / 其他

缺失的必填字段(客户姓名、投资者类型)在四问后补充提问。

输出

全部必填字段齐备后生成画像表。不要等 RM 确认,直接输出 YAML 并衔接下游。

profile_id: "P-{YYYYMMDD}-{3位序号}"
rm_name: "{客户经理姓名}"
client_name: "{客户姓名}"
risk_level: "R3"
amount: 500000
horizon: "中期"
goal: "资产增值"
liquidity: "低"
constraints: []
investor_type: "普通投资者"
mode: "A"
path_id: "WP-REC-01"
engine_reading_order:
  - engine/product-taxonomy.md
  - engine/filter-rules.md
  - engine/scoring-framework.md
  - engine/suitability-rules.md
  - engine/dual-track-methodology.md
quality_gates:
  - "风险等级匹配 (engine/suitability-rules.md §风险匹配矩阵)"
  - "产品禁售过滤 (engine/filter-rules.md §过滤规则详情)"
  - "评分置信度 (engine/scoring-framework.md §置信度规则)"
profile_completeness:
  mandatory_filled: "{已填数}/{总数}"
  missing_fields: []
  confidence: "高"
notes: ""

链接

下游技能:recommend-engine — 画像表生成后立即移交,不需要 RM 确认。

护栏

  • 风险等级仅使用 R1-R5,不得自创等级名称
  • 客户姓名和客户经理姓名为必填项
  • profile_id 格式为 P-YYYYMMDD-NNN
  • 预检不会因为部分字段缺失而报错——缺失是正常的,提问即可
  • 不要为了"快速完成"而跳步或合并问题

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,144. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.