business-intelligence by borghei/claude-skills
npx skills add https://github.com/borghei/claude-skills --skill business-intelligence为数据驱动决策提供专家级商业智能。
DATA SOURCES → ETL/ELT → DATA WAREHOUSE → SEMANTIC LAYER → DASHBOARDS
│ │ │ │ │
▼ ▼ ▼ ▼ ▼
CRM, ERP Transform Star Schema Metrics Def Tableau/PBI
APIs, DBs Clean, Load Fact/Dims Calculations Looker/etc
PRESENTATION LAYER
├── Executive dashboards
├── Operational reports
├── Self-service exploration
└── Embedded analytics
SEMANTIC LAYER
├── Business metrics definitions
├── Calculated fields
├── Hierarchies
└── Row-level security
DATA LAYER
├── Data warehouse (Snowflake/BigQuery/Redshift)
├── Data marts
├── Materialized views
└── Cached datasets
广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
高管仪表板:
┌─────────────────────────────────────────────────────────────┐
│ EXECUTIVE SUMMARY │
├─────────────────────────────────────────────────────────────┤
│ Revenue Pipeline Customers NPS │
│ $12.4M $45.2M 2,847 72 │
│ +15% YoY +22% QoQ +340 MTD +5 pts │
├─────────────────────────────────────────────────────────────┤
│ REVENUE TREND │ REVENUE BY SEGMENT │
│ [Line chart: 12 months] │ [Pie chart: segments] │
├────────────────────────────────┼─────────────────────────────┤
│ TOP ACCOUNTS │ KEY METRICS STATUS │
│ [Table: top 10] │ [KPI cards with RAG] │
└─────────────────────────────────────────────────────────────┘
运营仪表板:
┌─────────────────────────────────────────────────────────────┐
│ DAILY OPERATIONS │
├─────────────────────────────────────────────────────────────┤
│ Orders Today Tickets Open Avg Response SLA Met │
│ 1,247 89 12 min 98.5% │
│ vs Avg: +8% vs Avg: -12% vs Target: ✓ vs Target: ✓ │
├─────────────────────────────────────────────────────────────┤
│ HOURLY VOLUME │ QUEUE STATUS │
│ [Area chart: 24h] │ [Stacked bar by team] │
├────────────────────────────────┼─────────────────────────────┤
│ ALERTS │ TEAM PERFORMANCE │
│ [Alert list with severity] │ [Table: agents + metrics] │
└─────────────────────────────────────────────────────────────┘
视觉层次:
颜色使用:
STATUS COLORS
├── Green (#28A745): Good/On Track
├── Yellow (#FFC107): Warning/At Risk
├── Red (#DC3545): Critical/Off Track
└── Gray (#6C757D): Neutral/No Status
BRAND COLORS
├── Primary: Use for emphasis
├── Secondary: Supporting elements
└── Accent: Highlights only
DATA COLORS
├── Sequential: Light → Dark for ranges
├── Diverging: Different hues for pos/neg
└── Categorical: Distinct colors per category
图表选择:
| 数据类型 | 最佳图表 |
|---|---|
| 随时间趋势 | 折线图,面积图 |
| 部分与整体 | 饼图,环形图,树状图 |
| 比较 | 条形图,柱状图 |
| 分布 | 直方图,箱线图 |
| 关系 | 散点图,气泡图 |
| 地理信息 | 地图,分级统计图 |
# KPI Definition: [Metric Name]
## Business Context
- Owner: [Department/Role]
- Purpose: [Why this metric matters]
- Strategic alignment: [Goal it supports]
## Definition
- Formula: [Calculation]
- Data source: [System/Table]
- Granularity: [Daily/Weekly/Monthly]
## Targets
- Target: [Value]
- Threshold (Yellow): [Value]
- Critical (Red): [Value]
## Dimensions
- Time: [Day/Week/Month/Quarter/Year]
- Segments: [By region, product, etc.]
## Caveats
- [Known limitations]
- [Data quality issues]
财务类:
| 指标 | 公式 | 频率 |
|---|---|---|
| 收入 | 已成交金额总和 | 每日 |
| 月度经常性收入 | 月度经常性收入 | 每月 |
| 毛利率 | (收入 - 销售成本) / 收入 | 每月 |
| 客户获取成本 | 销售与营销支出 / 新客户数 | 每月 |
| 客户终身价值 | 每用户平均收入 × 毛利率 × 生命周期 | 每季度 |
客户类:
| 指标 | 公式 | 频率 |
|---|---|---|
| 活跃用户 | 日活,周活,月活 | 每日 |
| 流失率 | 流失客户数 / 总客户数 | 每月 |
| 净推荐值 | 推荐者 - 贬损者 | 每季度 |
| 客户满意度 | 满意数 / 总回复数 | 每周 |
运营类:
| 指标 | 公式 | 频率 |
|---|---|---|
| 吞吐量 | 单位数 / 时间 | 每小时 |
| 错误率 | 错误数 / 总数 | 每日 |
| 周期时间 | 结束时间 - 开始时间 | 每日 |
| 利用率 | 活跃数 / 容量 | 每日 |
计划报告:
report:
name: Weekly Sales Report
schedule: "0 8 * * MON" # Every Monday 8am
recipients:
- sales-team@company.com
- leadership@company.com
format: PDF
pages:
- Executive Summary
- Pipeline Analysis
- Rep Performance
- Forecast
阈值警报:
alert:
name: Revenue Below Target
metric: daily_revenue
condition: actual < target * 0.9
frequency: daily
channels:
- email: finance@company.com
- slack: #revenue-alerts
message: |
Daily revenue of ${actual} is ${pct_diff}% below target.
Top contributing factors: ${top_factors}
def generate_report(report_config):
"""
Automated report generation workflow
"""
# 1. Refresh data
refresh_data_sources(report_config['sources'])
# 2. Calculate metrics
metrics = calculate_metrics(report_config['metrics'])
# 3. Generate visualizations
charts = create_visualizations(metrics, report_config['charts'])
# 4. Build report
report = compile_report(
metrics=metrics,
charts=charts,
template=report_config['template']
)
# 5. Distribute
distribute_report(
report=report,
recipients=report_config['recipients'],
format=report_config['format']
)
return report
SELF-SERVICE MATURITY MODEL
Level 1: Report Consumers
├── View existing dashboards
├── Apply filters
└── Export data
Level 2: Data Explorers
├── Ad-hoc queries
├── Create simple charts
└── Share findings
Level 3: Report Builders
├── Design dashboards
├── Combine data sources
└── Create calculated fields
Level 4: Data Modelers
├── Create data models
├── Define metrics
└── Optimize performance
# Data Catalog Entry
## Dataset: sales_opportunities
### Description
Contains all sales opportunities from CRM
### Schema
| Column | Type | Description |
|--------|------|-------------|
| opp_id | STRING | Unique identifier |
| account_id | STRING | Related account |
| amount | DECIMAL | Deal value |
| stage | STRING | Pipeline stage |
| close_date | DATE | Expected close |
| owner_id | STRING | Sales rep |
### Refresh
- Frequency: Every 4 hours
- Source: Salesforce API
- Last refresh: 2024-01-15 08:00 UTC
### Usage Notes
- Filter by is_deleted = false
- Amount is always in USD
- Stage values: Prospect, Discovery, Demo, Proposal, Negotiation, Closed Won, Closed Lost
### Related Datasets
- accounts
- sales_reps
- products
SITUATION → COMPLICATION → RESOLUTION
1. SITUATION (Context)
"Last quarter, we set a goal to increase customer retention by 10%"
2. COMPLICATION (Problem/Opportunity)
"However, churn increased by 5% in our enterprise segment"
3. RESOLUTION (Insight + Action)
"Analysis shows onboarding time correlates with churn.
Reducing onboarding from 30 to 14 days could save $2M annually"
# Insight: [Title]
## What happened?
[Describe the observation in data]
## Why does it matter?
[Business impact and context]
## Why did it happen?
[Root cause analysis]
## What should we do?
[Recommended actions]
## Supporting Data
[Charts and metrics]
EXECUTIVE PRESENTATION STRUCTURE
1. Headlines First (2-3 key takeaways)
2. Context (why we're looking at this)
3. Key Findings (data + insights)
4. Implications (what it means)
5. Recommendations (what to do)
6. Appendix (detailed data)
仪表板性能:
OPTIMIZATION CHECKLIST
□ Limit visualizations per page (5-8 max)
□ Use data extracts vs live connections
□ Minimize calculated fields in viz
□ Use context filters effectively
□ Aggregate data at source when possible
□ Schedule refreshes during off-peak
□ Monitor query execution times
查询优化:
-- Bad: Full table scan
SELECT * FROM large_table
WHERE date >= '2024-01-01';
-- Good: Partitioned and filtered
SELECT required_columns
FROM large_table
WHERE partition_date >= '2024-01-01'
AND status = 'active'
LIMIT 10000;
访问控制:
security_model:
row_level_security:
- rule: region_access
filter: "region = user.region"
- rule: team_access
filter: "team_id IN user.teams"
object_permissions:
- role: viewer
permissions: [view, export]
- role: editor
permissions: [view, export, edit]
- role: admin
permissions: [view, export, edit, delete, publish]
数据质量监控:
DATA QUALITY CHECKS
├── Freshness: Is data current?
├── Completeness: Are all records present?
├── Accuracy: Do values make sense?
├── Consistency: Do related metrics align?
└── Uniqueness: Are there duplicates?
references/dashboard_patterns.md - 仪表板设计模式references/visualization_guide.md - 图表选择指南references/kpi_library.md - 标准 KPI 定义references/storytelling.md - 数据叙事技巧# Dashboard performance analyzer
python scripts/dashboard_analyzer.py --dashboard "Sales Overview"
# KPI calculator
python scripts/kpi_calculator.py --config metrics.yaml --output report.json
# Report generator
python scripts/report_generator.py --template weekly_sales --format pdf
# Data quality checker
python scripts/data_quality.py --dataset sales_opportunities --checks all
每周安装数
347
代码仓库
GitHub 星标数
29
首次出现
Jan 24, 2026
安全审计
安装于
opencode303
gemini-cli283
codex277
github-copilot262
cursor260
kimi-cli234
Expert-level business intelligence for data-driven decisions.
DATA SOURCES → ETL/ELT → DATA WAREHOUSE → SEMANTIC LAYER → DASHBOARDS
│ │ │ │ │
▼ ▼ ▼ ▼ ▼
CRM, ERP Transform Star Schema Metrics Def Tableau/PBI
APIs, DBs Clean, Load Fact/Dims Calculations Looker/etc
PRESENTATION LAYER
├── Executive dashboards
├── Operational reports
├── Self-service exploration
└── Embedded analytics
SEMANTIC LAYER
├── Business metrics definitions
├── Calculated fields
├── Hierarchies
└── Row-level security
DATA LAYER
├── Data warehouse (Snowflake/BigQuery/Redshift)
├── Data marts
├── Materialized views
└── Cached datasets
Executive Dashboard:
┌─────────────────────────────────────────────────────────────┐
│ EXECUTIVE SUMMARY │
├─────────────────────────────────────────────────────────────┤
│ Revenue Pipeline Customers NPS │
│ $12.4M $45.2M 2,847 72 │
│ +15% YoY +22% QoQ +340 MTD +5 pts │
├─────────────────────────────────────────────────────────────┤
│ REVENUE TREND │ REVENUE BY SEGMENT │
│ [Line chart: 12 months] │ [Pie chart: segments] │
├────────────────────────────────┼─────────────────────────────┤
│ TOP ACCOUNTS │ KEY METRICS STATUS │
│ [Table: top 10] │ [KPI cards with RAG] │
└─────────────────────────────────────────────────────────────┘
Operational Dashboard:
┌─────────────────────────────────────────────────────────────┐
│ DAILY OPERATIONS │
├─────────────────────────────────────────────────────────────┤
│ Orders Today Tickets Open Avg Response SLA Met │
│ 1,247 89 12 min 98.5% │
│ vs Avg: +8% vs Avg: -12% vs Target: ✓ vs Target: ✓ │
├─────────────────────────────────────────────────────────────┤
│ HOURLY VOLUME │ QUEUE STATUS │
│ [Area chart: 24h] │ [Stacked bar by team] │
├────────────────────────────────┼─────────────────────────────┤
│ ALERTS │ TEAM PERFORMANCE │
│ [Alert list with severity] │ [Table: agents + metrics] │
└─────────────────────────────────────────────────────────────┘
Visual Hierarchy:
Color Usage:
STATUS COLORS
├── Green (#28A745): Good/On Track
├── Yellow (#FFC107): Warning/At Risk
├── Red (#DC3545): Critical/Off Track
└── Gray (#6C757D): Neutral/No Status
BRAND COLORS
├── Primary: Use for emphasis
├── Secondary: Supporting elements
└── Accent: Highlights only
DATA COLORS
├── Sequential: Light → Dark for ranges
├── Diverging: Different hues for pos/neg
└── Categorical: Distinct colors per category
Chart Selection:
| Data Type | Best Charts |
|---|---|
| Trend over time | Line, Area |
| Part of whole | Pie, Donut, Treemap |
| Comparison | Bar, Column |
| Distribution | Histogram, Box Plot |
| Relationship | Scatter, Bubble |
| Geographic | Map, Choropleth |
# KPI Definition: [Metric Name]
## Business Context
- Owner: [Department/Role]
- Purpose: [Why this metric matters]
- Strategic alignment: [Goal it supports]
## Definition
- Formula: [Calculation]
- Data source: [System/Table]
- Granularity: [Daily/Weekly/Monthly]
## Targets
- Target: [Value]
- Threshold (Yellow): [Value]
- Critical (Red): [Value]
## Dimensions
- Time: [Day/Week/Month/Quarter/Year]
- Segments: [By region, product, etc.]
## Caveats
- [Known limitations]
- [Data quality issues]
Financial:
| Metric | Formula | Frequency |
|---|---|---|
| Revenue | Sum of closed won | Daily |
| MRR | Monthly recurring | Monthly |
| Gross Margin | (Rev - COGS) / Rev | Monthly |
| CAC | S&M Spend / New Customers | Monthly |
| LTV | ARPU × Margin × Lifetime | Quarterly |
Customer:
| Metric | Formula | Frequency |
|---|---|---|
| Active Users | DAU, WAU, MAU | Daily |
| Churn Rate | Lost / Total | Monthly |
| NPS | Promoters - Detractors | Quarterly |
| CSAT | Satisfied / Responses | Weekly |
Operations:
| Metric | Formula | Frequency |
|---|---|---|
| Throughput | Units / Time | Hourly |
| Error Rate | Errors / Total | Daily |
| Cycle Time | End - Start | Daily |
| Utilization | Active / Capacity | Daily |
Scheduled Reports:
report:
name: Weekly Sales Report
schedule: "0 8 * * MON" # Every Monday 8am
recipients:
- sales-team@company.com
- leadership@company.com
format: PDF
pages:
- Executive Summary
- Pipeline Analysis
- Rep Performance
- Forecast
Threshold Alerts:
alert:
name: Revenue Below Target
metric: daily_revenue
condition: actual < target * 0.9
frequency: daily
channels:
- email: finance@company.com
- slack: #revenue-alerts
message: |
Daily revenue of ${actual} is ${pct_diff}% below target.
Top contributing factors: ${top_factors}
def generate_report(report_config):
"""
Automated report generation workflow
"""
# 1. Refresh data
refresh_data_sources(report_config['sources'])
# 2. Calculate metrics
metrics = calculate_metrics(report_config['metrics'])
# 3. Generate visualizations
charts = create_visualizations(metrics, report_config['charts'])
# 4. Build report
report = compile_report(
metrics=metrics,
charts=charts,
template=report_config['template']
)
# 5. Distribute
distribute_report(
report=report,
recipients=report_config['recipients'],
format=report_config['format']
)
return report
SELF-SERVICE MATURITY MODEL
Level 1: Report Consumers
├── View existing dashboards
├── Apply filters
└── Export data
Level 2: Data Explorers
├── Ad-hoc queries
├── Create simple charts
└── Share findings
Level 3: Report Builders
├── Design dashboards
├── Combine data sources
└── Create calculated fields
Level 4: Data Modelers
├── Create data models
├── Define metrics
└── Optimize performance
# Data Catalog Entry
## Dataset: sales_opportunities
### Description
Contains all sales opportunities from CRM
### Schema
| Column | Type | Description |
|--------|------|-------------|
| opp_id | STRING | Unique identifier |
| account_id | STRING | Related account |
| amount | DECIMAL | Deal value |
| stage | STRING | Pipeline stage |
| close_date | DATE | Expected close |
| owner_id | STRING | Sales rep |
### Refresh
- Frequency: Every 4 hours
- Source: Salesforce API
- Last refresh: 2024-01-15 08:00 UTC
### Usage Notes
- Filter by is_deleted = false
- Amount is always in USD
- Stage values: Prospect, Discovery, Demo, Proposal, Negotiation, Closed Won, Closed Lost
### Related Datasets
- accounts
- sales_reps
- products
SITUATION → COMPLICATION → RESOLUTION
1. SITUATION (Context)
"Last quarter, we set a goal to increase customer retention by 10%"
2. COMPLICATION (Problem/Opportunity)
"However, churn increased by 5% in our enterprise segment"
3. RESOLUTION (Insight + Action)
"Analysis shows onboarding time correlates with churn.
Reducing onboarding from 30 to 14 days could save $2M annually"
# Insight: [Title]
## What happened?
[Describe the observation in data]
## Why does it matter?
[Business impact and context]
## Why did it happen?
[Root cause analysis]
## What should we do?
[Recommended actions]
## Supporting Data
[Charts and metrics]
EXECUTIVE PRESENTATION STRUCTURE
1. Headlines First (2-3 key takeaways)
2. Context (why we're looking at this)
3. Key Findings (data + insights)
4. Implications (what it means)
5. Recommendations (what to do)
6. Appendix (detailed data)
Dashboard Performance:
OPTIMIZATION CHECKLIST
□ Limit visualizations per page (5-8 max)
□ Use data extracts vs live connections
□ Minimize calculated fields in viz
□ Use context filters effectively
□ Aggregate data at source when possible
□ Schedule refreshes during off-peak
□ Monitor query execution times
Query Optimization:
-- Bad: Full table scan
SELECT * FROM large_table
WHERE date >= '2024-01-01';
-- Good: Partitioned and filtered
SELECT required_columns
FROM large_table
WHERE partition_date >= '2024-01-01'
AND status = 'active'
LIMIT 10000;
Access Control:
security_model:
row_level_security:
- rule: region_access
filter: "region = user.region"
- rule: team_access
filter: "team_id IN user.teams"
object_permissions:
- role: viewer
permissions: [view, export]
- role: editor
permissions: [view, export, edit]
- role: admin
permissions: [view, export, edit, delete, publish]
Data Quality Monitoring:
DATA QUALITY CHECKS
├── Freshness: Is data current?
├── Completeness: Are all records present?
├── Accuracy: Do values make sense?
├── Consistency: Do related metrics align?
└── Uniqueness: Are there duplicates?
references/dashboard_patterns.md - Dashboard design patternsreferences/visualization_guide.md - Chart selection guidereferences/kpi_library.md - Standard KPI definitionsreferences/storytelling.md - Data storytelling techniques# Dashboard performance analyzer
python scripts/dashboard_analyzer.py --dashboard "Sales Overview"
# KPI calculator
python scripts/kpi_calculator.py --config metrics.yaml --output report.json
# Report generator
python scripts/report_generator.py --template weekly_sales --format pdf
# Data quality checker
python scripts/data_quality.py --dataset sales_opportunities --checks all
Weekly Installs
347
Repository
GitHub Stars
29
First Seen
Jan 24, 2026
Security Audits
Gen Agent Trust HubPassSocketPassSnykPass
Installed on
opencode303
gemini-cli283
codex277
github-copilot262
cursor260
kimi-cli234
DOCX文件创建、编辑与分析完整指南 - 使用docx-js、Pandoc和Python脚本
42,600 周安装
Claude智能体开发指南:创建自主AI助手,掌握agent-development核心技巧
839 周安装
LobeHub Zustand 状态管理指南:架构模式、乐观更新与最佳实践
808 周安装
Trigger.dev配置指南:trigger.config.ts详解与构建扩展配置教程
831 周安装
高级数据工程师技能:生产级AI/ML数据系统架构、MLOps与性能优化
801 周安装
tRPC 端到端类型安全指南 - TypeScript API 开发与 React/Next.js 集成
794 周安装
React extract-errors 工具:自动提取和分配错误代码,提升开发效率
788 周安装