[q] Reduce verbosity of Copilot Agent Analysis reports

[Copilot]

Problem

The daily Copilot Agent Analysis workflow was generating extremely verbose reports (~200-300 lines) when 15-20 lines would suffice for daily monitoring. This was raised by @pelikhan in discussion #1993 with feedback to "reduce the verbosity of these reports."

Solution

Applied changes from Q workflow run 18636565580 to dramatically reduce report verbosity while maintaining essential performance metrics and actionable insights.

Key Changes

1. Simplified Mission & Success Criteria

• Added explicit goal: "Focus on concise summaries"
• Updated success criteria to target ~15-20 lines of essential information
• Added reminder: "Less is more"

2. Streamlined Summary Generation (Phase 3)

• Removed detailed 8-column PR summary table
• Changed to brief summary showing only:
  • Total PRs with success rate
  • PRs with issues (failures, closures)
  • Omit table unless notable PRs to highlight

3. Reduced Trend Analysis (Phase 4)

• Data collection: 3 days (was 7-30 days)
• Change threshold: >10% (was 5%)
• Removed: Weekly and monthly summary sections
• Focus: Only significant changes, skip stable metrics

4. Removed Instruction Changes Analysis (Phase 5)

• Eliminated entire section correlating instruction file changes with performance
• Rationale: Added unnecessary verbosity without clear value

5. Concise Discussion Template (Phase 6)

Before (~200+ lines):

• Detailed PR Summary Table
• Today's Performance (7 bullet points)
• Weekly Summary (7-day breakdown table)
• Monthly Summary (weekly trends, 30-day data)
• Instruction File Changes (full correlation analysis)
• Detailed PR Analysis (individual breakdowns for all PRs)
• Recommendations (large conditional section)
• Notes (additional context)

After (~15-20 lines):

# 🤖 Copilot Agent PR Analysis - [DATE]

## Summary
**Total PRs**: [count] | **Merged**: [count] ([%]%) | **Avg Duration**: [time]

## Performance Metrics
[3-day table only]

**Trend**: [Only if significant change >10%]

## Key Insights
[1-2 bullet points only]

---
_Generated by Copilot Agent Analysis (Run: [run_id])_

6. Updated Cache Management

• Retention: 90 days (was 365 days)
• Collection: 3 days maximum (was 7-30 days)

Impact

Report Length

• Before: ~200-300 lines with extensive tables and breakdowns
• After: ~15-25 lines with essential metrics only
• Reduction: ~90%

User Experience

• Reading time: Reduced from 5-10 minutes to 30-60 seconds
• Signal-to-noise: Much clearer action items, removed repetitive stable metrics
• Cognitive load: Immediate visibility of issues and notable PRs

Processing Efficiency

• Data collection: ~70% reduction (3 days vs 7-30 days)
• Token usage: Significantly reduced per run
• Analysis speed: Faster with smaller dataset

Validation

• ✅ Workflow compiles successfully (193.2 KB output)
• ✅ Build passes
• ✅ All tests pass (unit + integration)
• ✅ Code review completed with no issues
• ✅ File reduced from 631 to 450 lines (29% reduction in workflow instructions)
• ✅ No breaking changes (cache structure unchanged, backward compatible)

Example Output Comparison

Before

# 🤖 Copilot Agent PR Analysis - 2025-10-19

## Summary
[3 lines]

## PR Summary Table
[Large 8-column table with 10+ rows]

## Metrics
### Today's Performance
[7 bullet points]

### Weekly Summary (Last 7 Days)
[Performance metrics + daily breakdown table]

### Monthly Summary (Last 30 Days)
[Performance metrics + weekly trends table + monthly trends]

### Historical Comparison
[3 comparison bullets]

## Instruction File Changes
[Full section with file details and correlation analysis]

## Detailed PR Analysis
### Merged PRs ✅
[10+ individual PR breakdowns]
### Closed PRs ❌
[Individual closures]
### Open PRs ⏳
[Individual open PRs]

## Recommendations
[Large section with conditional recommendations]

## Notes
[Additional context]

---
[Verbose footer]

Total: ~200+ lines

After

# 🤖 Copilot Agent PR Analysis - 2025-10-19

## Summary
**Total PRs**: 10 | **Merged**: 8 (89%) | **Avg Duration**: 95m

## Performance Metrics
| Date | PRs | Merged | Success Rate | Avg Duration | Avg Comments |
|------|-----|--------|--------------|--------------|--------------|
| Oct 19 | 10 | 8 | 89% | 95m | 3.1 |
| Oct 18 | 5 | 4 | 80% | 25m | 0.4 |
| Oct 17 | 9 | 8 | 89% | 37m | 1.44 |

**Trend**: Duration increased 📉 (95m vs 31m avg, +207%)

## Key Insights
- Success rate stable at 89%
- Duration spike today due to complex features ([#1972](url), [#1963](url))

---
_Generated by Copilot Agent Analysis (Run: 18635640300)_

Total: ~18 lines

References

• Issue: [copilot-agent-analysis] Daily Copilot Agent Analysis - 2025-10-19 #1993
• Requester: @pelikhan
• Patch Source: Workflow run 18636565580

Original prompt

---

This section details on the original issue you should resolve

<issue_title>[q] Reduce verbosity of Copilot Agent Analysis reports</issue_title>
<issue_description># Q Workflow Optimization Report

Context

This PR addresses feedback from @pelikhan in discussion #1993 requesting to "reduce the verbosity of these reports" for the daily Copilot Agent Analysis workflow.

Issues Found (from live data)

Daily Copilot Agent Analysis Report (Discussion #1993)

• Observation: Current reports are extremely verbose with:
  • Extensive PR summary tables listing every PR
  • Detailed 3-day, 7-day, and monthly trend breakdowns
  • Individual PR analysis for every merged/closed/open PR
  • Instruction file correlation analysis
  • Verbose methodology and recommendations sections
• User Feedback: @pelikhan requested reduced verbosity
• Assessment: Reports contain ~200+ lines when 15-20 lines would suffice for daily monitoring

Changes Made

.github/workflows/copilot-agent-analysis.md

1. Simplified Mission Statement

• Added explicit goal: "Focus on concise summaries"
• Removed promise of weekly and monthly summaries

2. Reduced Summary Table (Phase 3)

• Changed from detailed PR-by-PR table to concise summary
• Only list PRs with issues (failures, closures)
• Omit detailed table unless notable PRs to highlight

3. Streamlined Trend Analysis (Phase 4.4)

• Changed from 7-day/30-day analysis to 3-day comparison
• Only report significant changes (>10% threshold)
• Skip stable metrics to reduce clutter
• Removed monthly summary requirement

4. Removed Instruction Changes Analysis (Phase 5)

• Eliminated entire section on instruction file correlation
• This was adding unnecessary verbosity without clear value

5. Concise Discussion Template (Phase 6)

• New template is ~15-20 lines vs previous ~100+ lines
• Removed: PR Summary Table, Detailed PR Analysis, Weekly/Monthly Summaries, Instruction Changes, Recommendations
• Kept: 3-day metrics table, Notable PRs (issues only), 1-2 Key Insights
• Simplified footer to single line

6. Updated Historical Data Collection

• Reduced from 7-30 days to 3 days maximum
• Simplified rebuild strategy
• Reduced cache retention from 365 days to 90 days

7. Stricter Trend Analysis Guidelines

• Minimum 3 days (was 7 days)
• Only report >10% changes (was 5%)
• Focus on significant changes only

8. Simplified Edge Cases

• Concise handling for no PRs, username changes, incomplete data
• 2-3 sentences max instead of verbose explanations

9. Updated Success Criteria

• Explicit goal: Keep discussion to ~15-20 lines
• Remove verbose tables and breakdowns
• Maximum 1-2 actionable insights
• Added reminder: "Less is more"

Expected Improvements

Report Length

• Before: ~200-300 lines with extensive tables and breakdowns
• After: ~15-25 lines with essential metrics only
• Reduction: ~90% shorter reports

Information Density

• Focus on actionable insights and significant changes only
• Remove repetitive daily data that doesn't show trends
• Eliminate speculative correlation analysis

Reader Experience

• Faster to scan and understand key metrics
• Immediate visibility of issues and notable PRs
• Less time spent reading stable/unchanged metrics

Processing Efficiency

• Reduced historical data collection (3 days vs 7-30 days)
• Faster analysis with smaller data set
• Lower token usage per run

Validation

Changes validated through:

• ✅ Manual review of workflow markdown structure
• ✅ Verification that essential metrics are preserved
• ✅ Confirmed removal of verbose sections
• ✅ Template simplified while maintaining clarity

Note: Workflow will be automatically compiled after merge. The .lock.yml file is intentionally not included in this PR.

Example Output Comparison

Before (Verbose)

# 🤖 Copilot Agent PR Analysis - 2025-10-19

## Summary
[3 lines]

## PR Summary Table
[Large table with 10+ rows and 8 columns]

## Metrics

### Today's Performance
[7 bullet points]

### Weekly Summary (Last 7 Days)
[Performance metrics, daily breakdown table with 7 rows]

### Monthly Summary (Last 30 Days)
[Performance metrics, weekly trends table, monthly trends bullets]

### Historical Comparison
[3 comparison bullets]

## Instruction File Changes
[Full section with file details and correlation analysis]

## Detailed PR Analysis

### Merged PRs ✅
[10+ individual PR breakdowns]

### Closed PRs ❌
[Individual closures]

### Open PRs ⏳
[Individual open PRs]

## Recommendations
[Large section with conditional recommendations]

## Notes
[Additional context]

---
[Verbose footer]

Total: ~200+ lines

After (Concise)

# 🤖 Copilot Agent PR Analysis - [DATE]

## Summary
**Total PRs**: 10 | **Merged**: 8 (89%) | **Avg Duration**: 95m

## Performance Metrics
| Date | PRs | Merged | Success Rate | Avg Duration | Avg Comments |
|------|-----|--------|-...

</details>

Fixes githubnext/gh-aw#1998

<!-- START COPILOT CODING AGENT TIPS -->
---

💬 Share your feedback on Copilot coding agent for the chance to win a $200 gift card! Click [here](https://survey3.medallia.com/?EAHeSx-AP01bZqG0Ld9QLQ) to start the survey.

[github-actions]

Agentic Changeset Generator triggered by this pull request.