[refactoring] Extract Python Chart Discussion Report Pattern into shared component

[github-actions]

Pattern Overview

The Python Chart Discussion Report pattern appears in 7 workflows that generate visualizations with Python/matplotlib, upload charts as assets, and create discussion reports with embedded chart images. Extracting this into a shared component would eliminate ~700 lines of duplicated code and ensure consistent report formatting.

Current Usage

This pattern appears in the following workflows:

• daily-issues-report.md (lines 31-34, 187-230, 235-262)
• daily-code-metrics.md (lines 30-31, 66-230, 240-365)
• python-data-charts.md
• daily-copilot-token-report.md
• daily-performance-summary.md
• session-analysis-charts.md (via shared import)
• trending-charts-simple.md (via shared import)

Proposed Shared Component

File: .github/workflows/shared/python-chart-discussion-report.md

Configuration:
```yaml

imports:

• shared/python-dataviz.md
• shared/reporting.md

safe-outputs:
upload-asset:
create-discussion:
max: 1
close-older-discussions: true

tools:
cache-memory: true

Python Chart Discussion Report Guide

This component extends python-dataviz.md for workflows that generate comprehensive reports with multiple visualization charts.

Standard Report Structure

All chart-based reports should follow this structure:

```markdown

[Report Title] - [Date]

Executive Summary

[2-3 paragraphs summarizing key findings, trends, and recommendations]

Key Statistics:

• Metric 1: [value]
• Metric 2: [value]
• Metric 3: [value]

Key Visualizations

[Chart 1 Title]

[2-3 sentence analysis of what this chart shows and key insights]

[Chart 2 Title]

[2-3 sentence analysis of what this chart shows and key insights]

[Chart 3 Title] (if applicable)

[2-3 sentence analysis]

📊 Detailed Metrics and Analysis

[Section 1]

[Detailed tables, metrics, and analysis]

Metric	Value	Change (7d)	Change (30d)
...	...	...	...

[Section 2]

[More detailed analysis]

💡 Insights and Recommendations

1. [Specific actionable recommendation based on data]
2. [Another recommendation]
3. [Additional recommendation]
4. [...]

---

Report generated by [Workflow Name] workflow
Data period: [period] | Generated: [timestamp]
```

Chart Upload Process

After generating charts in /tmp/gh-aw/python/charts/:

Step 1: Upload charts as assets
```bash

Upload each chart and collect URLs

declare -A chart_urls

for chart_file in /tmp/gh-aw/python/charts/*.png; do
if [ -f "$chart_file" ]; then
chart_name=$(basename "$chart_file" .png)
echo "Uploading $chart_name..."

# Use upload-asset tool (returns URL)
# Store URL in associative array
# chart_urls["$chart_name"]="$returned_url"

fi
done
```

Step 2: Create discussion with embedded charts

Use the collected URLs to embed charts in your discussion markdown. Follow the standard report structure above.

Standard Chart Set

Most workflows should generate 2-4 charts:

1. Primary Trend Chart: Time series showing main metrics over 30 days

   • Line chart with multiple series
   • Include 7-day moving average
   • Annotate significant changes

2. Distribution Chart: Breakdown of current state

   • Bar chart or pie chart
   • Show proportions or counts
   • Sort by significance

3. Comparison Chart (optional): Before/after or comparative analysis

   • Grouped bars or side-by-side comparison
   • Highlight differences
   • Include percentage changes

4. Detailed Analysis Chart (optional): Deep dive into specific aspect

   • Heatmap, scatter plot, or specialized visualization
   • Focus on specific insight

Chart Quality Standards

All charts MUST meet these requirements (from python-dataviz.md):

• DPI: 300 minimum for publication quality
• Figure Size: 12x7 inches (or 10x6 for smaller charts)
• Styling: Use seaborn styling (sns.set_style("whitegrid"))
• Color Palette: Professional colors (sns.set_palette("husl") or custom)
• Labels: Clear titles, axis labels, and legends
• Grid Lines: Enable for readability (ax.grid(True, alpha=0.3))
• Save Format: PNG with bbox_inches='tight'

Chart Generation Template

```python
#!/usr/bin/env python3
"""
Chart generation script for [workflow name]
Generates [number] charts for the daily report
"""
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from datetime import datetime

Set consistent style

sns.set_style("whitegrid")
sns.set_palette("husl")

Load data (NEVER inline - always from external file)

data = pd.read_json('/tmp/gh-aw/python/data/metrics.json')

Chart 1: Trend Chart

def generate_trend_chart():
fig, ax = plt.subplots(figsize=(12, 7), dpi=300)

# Your plotting code
data.plot(x='date', y=['metric1', 'metric2'], ax=ax)

# Customize
ax.set_title('Metric Trends - Last 30 Days', fontsize=16, fontweight='bold')
ax.set_xlabel('Date', fontsize=12)
ax.set_ylabel('Value', fontsize=12)
ax.legend(loc='best')
ax.grid(True, alpha=0.3)

# Save
plt.savefig('/tmp/gh-aw/python/charts/trend_chart.png',
            dpi=300, bbox_inches='tight', facecolor='white')
print("✓ Trend chart generated")

Chart 2: Distribution Chart

def generate_distribution_chart():
fig, ax = plt.subplots(figsize=(10, 6), dpi=300)

# Your plotting code
data['category'].value_counts().plot(kind='barh', ax=ax)

# Customize
ax.set_title('Distribution by Category', fontsize=16, fontweight='bold')
ax.set_xlabel('Count', fontsize=12)
ax.grid(True, alpha=0.3)

# Save
plt.savefig('/tmp/gh-aw/python/charts/distribution_chart.png',
            dpi=300, bbox_inches='tight', facecolor='white')
print("✓ Distribution chart generated")

Generate all charts

if name == 'main':
generate_trend_chart()
generate_distribution_chart()
print("All charts generated successfully")
```

Discussion Creation Best Practices

1. Title format: [Category] Report Name - YYYY-MM-DD
2. Executive summary: 2-3 paragraphs max, highlight key findings
3. Chart placement: Embed charts early, before detailed metrics
4. Analysis text: 2-3 sentences per chart explaining insights
5. Collapsible details: Put detailed tables/metrics in <details> tag
6. Recommendations: 3-5 bullet points, specific and actionable
7. Metadata footer: Include workflow name, data period, generation timestamp

Cache Memory Integration

Store reusable chart generation code in cache:

```bash

Check for cached chart utilities

if [ -f /tmp/gh-aw/cache-memory/chart_utils.py ]; then
cp /tmp/gh-aw/cache-memory/chart_utils.py /tmp/gh-aw/python/
echo "Using cached chart utilities"
fi

Save new utilities to cache

if [ -f /tmp/gh-aw/python/chart_utils.py ]; then
cp /tmp/gh-aw/python/chart_utils.py /tmp/gh-aw/cache-memory/
echo "Saved chart utilities to cache"
fi
```

Complete Example Usage

```yaml
imports:

• shared/python-chart-discussion-report.md
• shared/issues-data-fetch.md # Or other data source

safe-outputs:
create-discussion:
category: "audits"
```

Then in your workflow:

1. Load and process data
2. Generate charts with Python
3. Upload charts as assets (collect URLs)
4. Create discussion using standard structure
5. Embed chart URLs in markdown
   ```

Usage Example:
```yaml

In a workflow

imports:

• shared/python-chart-discussion-report.md
• shared/issues-data-fetch.md
  ```

Impact

• Workflows affected: 7 workflows
• Lines saved: ~700 lines (100 per workflow)
• Maintenance benefit:
  • Consistent report formatting across all chart-based workflows
  • Enforces chart quality standards (DPI, sizing, styling)
  • Easier to update report structure (cascades to all workflows)
  • Reduces copy-paste errors in chart generation code

Implementation Plan

1. Create shared component at .github/workflows/shared/python-chart-discussion-report.md
2. Pilot: Update daily-issues-report.md to use shared component
3. Validate pilot: Check chart quality and discussion format
4. Update daily-code-metrics.md
5. Update python-data-charts.md
6. Update remaining 4 workflows
7. Test all workflows to ensure charts display correctly
8. Update documentation with examples

Related Analysis

This recommendation comes from the Workflow Pattern Harvester analysis run on 2026-01-16.

AI generated by Workflow Pattern Harvester