← Back to Skills Marketplace
nimolc

SkillGapDiagnosis(业绩缺口分析)

by NemoTech · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
33
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install skill-gap-diagnosis
Description
業務缺口診斷 Skill。整合業績、競賽、榮譽、續保率、客戶增長、技能短板等多維數據,統一換算為保障保費口徑,計算缺口金額、需成交件數、緊急度/容易度標籤,按優先級排序,輸出多維度診斷結論。不輸出行動建議。在用戶表達「診斷」、「分析」、「整體情況」、「缺口」、「進度」、「還差多少」或同時詢問多個維度業績時觸發。禁...
README (SKILL.md)

業務缺口診斷

核心职责

整合業績、競賽、榮譽、續保率、客戶增長、技能短板等多維數據,統一換算為保障保費口徑,計算缺口金額、需成交件數、緊急度/容易度標籤,按優先級排序,輸出多維度診斷結論。

約束:本 Skill 只輸出診斷結論,不輸出行動建議。行動建議由 skill_action_suggestion 統一生成。

觸發條件

正向觸發(滿足任一)

  • 用戶明確表達「診斷」、「分析」、「整體情況」、「缺口」、「進度」、「還差多少」
  • 用戶同時詢問多個維度業績(如競賽+榮譽+考核)
  • 用戶說「我這個月業績怎麼樣」、「MDRT 進度如何」、「幫我看看缺口」

禁止觸發

  • 用戶只問單一指標:「FYC 多少」、「保費多少」、「件數多少」 → 直接調對應 Tool
  • 用戶只問單一排名:「競賽排名第幾」 → 直接調 Data_Campaign
  • 閒聊/問候:「你好」、「在嗎」 → Agent 直接回復
  • 意圖不明且未澄清:「幫我看看」 → Agent 反問澄清

執行流程

步驟 1:並行調用多維數據接口

並行調用以下接口,每個接口超時 10s,失敗時跳過該維度:

接口 參數 失敗策略
Data_Performance dataType=ALL 缺失業績基礎數據
Data_Campaign queryMode=FULL, hasGapOnly=true 缺失競賽缺口
Data_Honor honorType=ALL 缺失榮譽進度
Data_Renewal policyType=ALL 缺失續保率
Data_CustomerGrowth 默認參數 缺失客戶增長
Data_Skill 默認參數 缺失技能短板
Data_Assessment(可選) assessmentType 按需 缺失考核數據
Data_NewAllowance(可選) 默認參數 缺失新人津貼

步驟 2:調用計算腳本

調用 scripts/compute_gaps.py,基於原始數據計算:

  • 缺口金額 = target - actual
  • 需成交件數 = ceil(缺口金額 / 件均保費)
  • 緊急度標籤:截止 ≤7 天為「緊急」,≤30 天為「重要」,其餘為「常規」
  • 容易度標籤:缺口 \x3C 日均產能 × 剩餘天數時為「容易」
  • 優先級排序:緊急度優先,其次按缺口金額降序

步驟 3:生成診斷文案

加載 references/diagnosis.md,將計算結果注入變量,生成診斷文案。

約束

  • 禁止機械編號(如「1. 2. 3.」列表)
  • 禁止暴露接口名稱(不出現 Data_Performance 等)
  • 金額保留原始單位(HKD)
  • 排名數據必須附加時效提示:「數據截至 {dataAsOfDate},由於排名會隨時間動態變化,請以實際達標情況為準」
  • 禁止承諾結果
  • daysRemaining > 60 不提及倒數天數

輸入參數

見 schema/input.json

輸出格式

見 schema/output.json

異常與降級

異常場景 處理策略 用戶感知
單個 API 超時 跳過該維度,缺失字段返回 null 「XX數據暫時無法獲取,以下為基於其他維度的分析」
全部 API 不可用 降級知識庫模式 「當前無法獲取實時數據,以下為通用建議」
件均保費缺失 requiredCases 返回 null 跳過件數建議,聚焦金額缺口
無缺口數據 gapIndicators 返回空數組 「恭喜,您當前無未達標缺口」
考核資格不符 該考核不納入 gapIndicators 不展示無資格考核

業務規則

  • 緊急度:截止 ≤7 天為「緊急」,≤30 天為「重要」,其餘為「常規」
  • 容易度:缺口 \x3C 日均產能 × 剩餘天數時為「容易」
  • 金額精度:保留兩位小數,單位 HKD
  • 排名數據必須提示時效性
  • daysRemaining > 60 不提及倒數天數
Usage Guidance
Install only if you are comfortable with the skill using broad insurance business performance data and with it sometimes producing recommendations, despite claiming to be diagnosis-only. It should be treated as a review-needed package until the publisher removes or clearly discloses the advice-generation behavior.
Capability Assessment
Purpose & Capability
The declared purpose is diagnostic-only analysis of insurance performance gaps, but the prompt examples, fallback text, and output fields include actionable recommendations and knowledge-retrieval hooks.
Instruction Scope
Trigger limits are clear, but the runtime instructions conflict: SKILL.md says action suggestions belong to another skill while diagnosis.md instructs the model to recommend follow-up, product mix, training, and higher goals.
Install Mechanism
The package is a normal skill artifact with markdown instructions, JSON schemas, tests, and one Python calculation script; metadata shows no dependencies and the static scan is clean.
Credentials
The requested data tools cover performance, campaigns, honors, renewal, customer growth, skills, assessments, and allowances, which is broad but coherent for multi-dimensional gap diagnosis.
Persistence & Privilege
No persistence, background worker, credential harvesting, privilege escalation, or hidden network behavior is evident; the script only reads optional input JSON files and can write a user-specified output file when run directly.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install skill-gap-diagnosis
  3. After installation, invoke the skill by name or use /skill-gap-diagnosis
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of skill-gap-diagnosis. - Integrates multi-dimensional data (performance, campaign, honors, renewal rate, customer growth, skill gaps) to calculate business shortfall. - Unifies metrics to protection premium caliber and computes gap amount, required cases, urgency/easiness tags, prioritizes, and outputs a diagnosis summary. - Only provides diagnosis conclusions; action suggestions are not included. - Triggered only for multi-dimensional or progress/shortfall analysis queries, not for single indicators, rankings, or general chat. - Robust error handling to inform users of data unavailability or fallback to general advice.
Metadata
Slug skill-gap-diagnosis
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is SkillGapDiagnosis(业绩缺口分析)?

業務缺口診斷 Skill。整合業績、競賽、榮譽、續保率、客戶增長、技能短板等多維數據,統一換算為保障保費口徑,計算缺口金額、需成交件數、緊急度/容易度標籤,按優先級排序,輸出多維度診斷結論。不輸出行動建議。在用戶表達「診斷」、「分析」、「整體情況」、「缺口」、「進度」、「還差多少」或同時詢問多個維度業績時觸發。禁... It is an AI Agent Skill for Claude Code / OpenClaw, with 33 downloads so far.

How do I install SkillGapDiagnosis(业绩缺口分析)?

Run "/install skill-gap-diagnosis" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is SkillGapDiagnosis(业绩缺口分析) free?

Yes, SkillGapDiagnosis(业绩缺口分析) is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does SkillGapDiagnosis(业绩缺口分析) support?

SkillGapDiagnosis(业绩缺口分析) is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created SkillGapDiagnosis(业绩缺口分析)?

It is built and maintained by NemoTech (@nimolc); the current version is v1.0.0.

💬 Comments