create-viz

Name: create-viz
Author: anthropics

4.1kinstalls2.7kforks

How to install create-viz?

anthropics/knowledge-work-plugins/create-viz

$npx -y skills add anthropics/knowledge-work-plugins --skill create-viz

Installs into the current project.

Summary

Create publication-quality visualizations with Python. Use when turning query results or a DataFrame into a chart, selecting the right chart type for a trend or comparison, generating a plot for a report or presentation, or needing an interactive chart with hover and zoom.

Files · 1

View on GitHub

SKILL.md

1	# /create-viz - Create Visualizations
2
3	> If you see unfamiliar placeholders or need to check which tools are connected, see [CONNECTORS.md](../../CONNECTORS.md).
4
5	Create publication-quality data visualizations using Python. Generates charts from data with best practices for clarity, accuracy, and design.
6
7	## Usage
8
9	```
10	/create-viz <data source> [chart type] [additional instructions]
11	```
12
13	## Workflow
14
15	### 1. Understand the Request
16
17	Determine:
18
19	- Data source: Query results, pasted data, CSV/Excel file, or data to be queried
20	- Chart type: Explicitly requested or needs to be recommended
21	- Purpose: Exploration, presentation, report, dashboard component
22	- Audience: Technical team, executives, external stakeholders
23
24	### 2. Get the Data
25
26	If data warehouse is connected and data needs querying:
27	1. Write and execute the query
28	2. Load results into a pandas DataFrame
29
30	If data is pasted or uploaded:
31	1. Parse the data into a pandas DataFrame
32	2. Clean and prepare as needed (type conversions, null handling)
33
34	If data is from a previous analysis in the conversation:
35	1. Reference the existing data
36
37	### 3. Select Chart Type
38
39	If the user didn't specify a chart type, recommend one based on the data and question:
40
41	\| Data Relationship \| Recommended Chart \|
42	\|---\|---\|
43	\| Trend over time \| Line chart \|
44	\| Comparison across categories \| Bar chart (horizontal if many categories) \|
45	\| Part-to-whole composition \| Stacked bar or area chart (avoid pie charts unless <6 categories) \|
46	\| Distribution of values \| Histogram or box plot \|
47	\| Correlation between two variables \| Scatter plot \|
48	\| Two-variable comparison over time \| Dual-axis line or grouped bar \|
49	\| Geographic data \| Choropleth map \|
50	\| Ranking \| Horizontal bar chart \|
51	\| Flow or process \| Sankey diagram \|
52	\| Matrix of relationships \| Heatmap \|
53
54	Explain the recommendation briefly if the user didn't specify.
55
56	### 4. Generate the Visualization
57
58	Write Python code using one of these libraries based on the need:
59
60	- matplotlib + seaborn: Best for static, publication-quality charts. Default choice.
61	- plotly: Best for interactive charts or when the user requests interactivity.
62
63	Code requirements:
64
65	```python
66	import matplotlib.pyplot as plt
67	import seaborn as sns
68	import pandas as pd
69
70	# Set professional style
71	plt.style.use('seaborn-v0_8-whitegrid')
72	sns.set_palette("husl")
73
74	# Create figure with appropriate size
75	fig, ax = plt.subplots(figsize=(10, 6))
76
77	# [chart-specific code]
78
79	# Always include:
80	ax.set_title('Clear, Descriptive Title', fontsize=14, fontweight='bold')
81	ax.set_xlabel('X-Axis Label', fontsize=11)
82	ax.set_ylabel('Y-Axis Label', fontsize=11)
83
84	# Format numbers appropriately
85	# - Percentages: '45.2%' not '0.452'
86	# - Currency: '$1.2M' not '1200000'
87	# - Large numbers: '2.3K' or '1.5M' not '2300' or '1500000'
88
89	# Remove chart junk
90	ax.spines['top'].set_visible(False)
91	ax.spines['right'].set_visible(False)
92
93	plt.tight_layout()
94	plt.savefig('chart_name.png', dpi=150, bbox_inches='tight')
95	plt.show()
96	```
97
98	### 5. Apply Design Best Practices
99
100	Color:
101	- Use a consistent, colorblind-friendly palette
102	- Use color meaningfully (not decoratively)
103	- Highlight the key data point or trend with a contrasting color
104	- Grey out less important reference data
105
106	Typography:
107	- Descriptive title that states the insight, not just the metric (e.g., "Revenue grew 23% YoY" not "Revenue by Month")
108	- Readable axis labels (not rotated 90 degrees if avoidable)
109	- Data labels on key points when they add clarity
110
111	Layout:
112	- Appropriate whitespace and margins
113	- Legend placement that doesn't obscure data
114	- Sorted categories by value (not alphabetically) unless there's a natural order
115
116	Accuracy:
117	- Y-axis starts at zero for bar charts
118	- No misleading axis breaks without clear notation
119	- Consistent scales when comparing panels
120	- Appropriate precision (don't show 10 decimal places)
121
122	### 6. Save and Present
123
124	1. Save the chart as a PNG file with descriptive name
125	2. Display the chart to the user
126	3. Provide the code used so they can modify it
127	4. Suggest variations (different chart type, different grouping, zoomed time range)
128
129	## Examples
130
131	```
132	/create-viz Show monthly revenue for the last 12 months as a line chart with the trend highlighted
133	```
134
135	```
136	/create-viz Here's our NPS data by product: [pastes data]. Create a horizontal bar chart ranking products by score.
137	```
138
139	```
140	/create-viz Query the orders table and create a heatmap of order volume by day-of-week and hour
141	```
142
143	## Tips
144
145	- If you want interactive charts (hover, zoom, filter), mention "interactive" and Claude will use plotly
146	- Specify "presentation" if you need larger fonts and higher contras

People who liked this also viewed

6 picks

Type

create-viz

4.1kinstalls2.7kforks

View on GitHub

How to install create-viz?

anthropics/knowledge-work-plugins/create-viz

$npx -y skills add anthropics/knowledge-work-plugins --skill create-viz

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# /create-viz - Create Visualizations
2
3	> If you see unfamiliar placeholders or need to check which tools are connected, see [CONNECTORS.md](../../CONNECTORS.md).
4
5	Create publication-quality data visualizations using Python. Generates charts from data with best practices for clarity, accuracy, and design.
6
7	## Usage
8
9	```
10	/create-viz <data source> [chart type] [additional instructions]
11	```
12
13	## Workflow
14
15	### 1. Understand the Request
16
17	Determine:
18
19	- Data source: Query results, pasted data, CSV/Excel file, or data to be queried
20	- Chart type: Explicitly requested or needs to be recommended
21	- Purpose: Exploration, presentation, report, dashboard component
22	- Audience: Technical team, executives, external stakeholders
23
24	### 2. Get the Data
25
26	If data warehouse is connected and data needs querying:
27	1. Write and execute the query
28	2. Load results into a pandas DataFrame
29
30	If data is pasted or uploaded:
31	1. Parse the data into a pandas DataFrame
32	2. Clean and prepare as needed (type conversions, null handling)
33
34	If data is from a previous analysis in the conversation:
35	1. Reference the existing data
36
37	### 3. Select Chart Type
38
39	If the user didn't specify a chart type, recommend one based on the data and question:
40
41	\| Data Relationship \| Recommended Chart \|
42	\|---\|---\|
43	\| Trend over time \| Line chart \|
44	\| Comparison across categories \| Bar chart (horizontal if many categories) \|
45	\| Part-to-whole composition \| Stacked bar or area chart (avoid pie charts unless <6 categories) \|
46	\| Distribution of values \| Histogram or box plot \|
47	\| Correlation between two variables \| Scatter plot \|
48	\| Two-variable comparison over time \| Dual-axis line or grouped bar \|
49	\| Geographic data \| Choropleth map \|
50	\| Ranking \| Horizontal bar chart \|
51	\| Flow or process \| Sankey diagram \|
52	\| Matrix of relationships \| Heatmap \|
53
54	Explain the recommendation briefly if the user didn't specify.
55
56	### 4. Generate the Visualization
57
58	Write Python code using one of these libraries based on the need:
59
60	- matplotlib + seaborn: Best for static, publication-quality charts. Default choice.
61	- plotly: Best for interactive charts or when the user requests interactivity.
62
63	Code requirements:
64
65	```python
66	import matplotlib.pyplot as plt
67	import seaborn as sns
68	import pandas as pd
69
70	# Set professional style
71	plt.style.use('seaborn-v0_8-whitegrid')
72	sns.set_palette("husl")
73
74	# Create figure with appropriate size
75	fig, ax = plt.subplots(figsize=(10, 6))
76
77	# [chart-specific code]
78
79	# Always include:
80	ax.set_title('Clear, Descriptive Title', fontsize=14, fontweight='bold')
81	ax.set_xlabel('X-Axis Label', fontsize=11)
82	ax.set_ylabel('Y-Axis Label', fontsize=11)
83
84	# Format numbers appropriately
85	# - Percentages: '45.2%' not '0.452'
86	# - Currency: '$1.2M' not '1200000'
87	# - Large numbers: '2.3K' or '1.5M' not '2300' or '1500000'
88
89	# Remove chart junk
90	ax.spines['top'].set_visible(False)
91	ax.spines['right'].set_visible(False)
92
93	plt.tight_layout()
94	plt.savefig('chart_name.png', dpi=150, bbox_inches='tight')
95	plt.show()
96	```
97
98	### 5. Apply Design Best Practices
99
100	Color:
101	- Use a consistent, colorblind-friendly palette
102	- Use color meaningfully (not decoratively)
103	- Highlight the key data point or trend with a contrasting color
104	- Grey out less important reference data
105
106	Typography:
107	- Descriptive title that states the insight, not just the metric (e.g., "Revenue grew 23% YoY" not "Revenue by Month")
108	- Readable axis labels (not rotated 90 degrees if avoidable)
109	- Data labels on key points when they add clarity
110
111	Layout:
112	- Appropriate whitespace and margins
113	- Legend placement that doesn't obscure data
114	- Sorted categories by value (not alphabetically) unless there's a natural order
115
116	Accuracy:
117	- Y-axis starts at zero for bar charts
118	- No misleading axis breaks without clear notation
119	- Consistent scales when comparing panels
120	- Appropriate precision (don't show 10 decimal places)
121
122	### 6. Save and Present
123
124	1. Save the chart as a PNG file with descriptive name
125	2. Display the chart to the user
126	3. Provide the code used so they can modify it
127	4. Suggest variations (different chart type, different grouping, zoomed time range)
128
129	## Examples
130
131	```
132	/create-viz Show monthly revenue for the last 12 months as a line chart with the trend highlighted
133	```
134
135	```
136	/create-viz Here's our NPS data by product: [pastes data]. Create a horizontal bar chart ranking products by score.
137	```
138
139	```
140	/create-viz Query the orders table and create a heatmap of order volume by day-of-week and hour
141	```
142
143	## Tips
144
145	- If you want interactive charts (hover, zoom, filter), mention "interactive" and Claude will use plotly
146	- Specify "presentation" if you need larger fonts and higher contras

People who liked this also viewed

6 picks

Type