Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
Use for: Trends over time
Example: Daily active users, revenue by month
When to use:
- Continuous data
- Show changes over time
- Compare multiple series
Comparison → Bar Chart
Use for: Comparing categories
Example: Revenue by product, sales by region
When to use:
- Categorical data
- Ranking (top 10)
- Part-to-whole (stacked bars)
Distribution → Histogram
Use for: Frequency distribution
Example: Order value distribution, user age ranges
When to use:
- Show data spread
- Identify outliers
- Normal vs skewed
Correlation → Scatter Plot
Use for: Relationship between variables
Example: Marketing spend vs revenue, price vs conversion
When to use:
- Two continuous variables
- Identify clusters
- Show outliers
Part-to-Whole → Pie Chart (Use Sparingly!)
Use for: Proportion of total (max 5 slices)
Example: Market share, traffic sources
Better alternative: Bar chart (easier to compare)
Geographic → Map
Use for: Location-based data
Example: Sales by state, user density
When to use:
- Spatial patterns
- Regional comparison
-- Metric: Monthly Recurring RevenueSELECT
DATE_TRUNC('month', subscription_start) ASmonth,
SUM(monthly_price) AS mrr,
COUNT(DISTINCT user_id) AS subscribers
FROM subscriptions
WHERE status ='active'AND subscription_start >= DATE_TRUNC('month', CURRENT_DATE-INTERVAL'12 months')
GROUPBY1ORDERBY1DESC-- Metric: Churn RateSELECT
DATE_TRUNC('month', cancelled_date) ASmonth,
COUNT(*) AS churned_customers,
ROUND(
COUNT(*)::NUMERIC/LAG(COUNT(*)) OVER (ORDERBY DATE_TRUNC('month', cancelled_date)) *100,
2
) AS churn_rate_pct
FROM subscriptions
WHERE status ='cancelled'AND cancelled_date >= DATE_TRUNC('month', CURRENT_DATE-INTERVAL'12 months')
GROUPBY1ORDERBY1DESC
Interactive Filters
Date Range Selector
-- Parameterized query in MetabaseSELECT
product_name,
SUM(revenue) AS total_revenue
FROM sales
WHERE sale_date BETWEEN {{start_date}} AND {{end_date}}
GROUPBY product_name
ORDERBY total_revenue DESC
LIMIT 10-- Parameters:-- start_date: Date field-- end_date: Date field
Multi-Select Filter
-- Filter by multiple regionsSELECT
region,
product_category,
SUM(revenue) AS revenue
FROM sales
WHERE region IN ({{regions}})
AND sale_date >=CURRENT_DATE-INTERVAL'30 days'GROUPBY region, product_category
-- Parameter:-- regions: Field filter on sales.region (multi-select)
Performance Optimization
Pre-Aggregation
-- Create materialized view for fast dashboard queriesCREATE MATERIALIZED VIEW daily_revenue_summary ASSELECTDATE(order_date) ASdate,
product_id,
region,
SUM(order_amount) AS revenue,
COUNT(*) AS order_count,
AVG(order_amount) AS avg_order_value
FROM orders
GROUPBY1, 2, 3;
-- Refresh nightlyCREATE INDEX ON daily_revenue_summary (date, region);
-- Query uses summary (fast)SELECT
region,
SUM(revenue) AS total_revenue
FROM daily_revenue_summary
WHEREdate>=CURRENT_DATE-INTERVAL'30 days'GROUPBY region;
Incremental Refresh
# Update only new dataimport pandas as pd
from datetime import datetime, timedelta
defincremental_refresh():
# Get last refresh timestamp
last_refresh = get_last_refresh_time()
# Query only new data
new_data = query_database(f"""
SELECT * FROM orders
WHERE updated_at > '{last_refresh}'
""")
# Append to existing data
append_to_dashboard_data(new_data)
# Update refresh timestamp
set_last_refresh_time(datetime.now())
Drill-Through & Drill-Down
Drill-Down (Hierarchy)
Revenue by Region
↓ (click region)
Revenue by Store
↓ (click store)
Revenue by Product
