Analytics & Reporting Guide

Track costs, performance, and ROI with built-in analytics dashboard.

Overview

ReviewRouter includes comprehensive analytics to help you:

Track costs across providers and over time
Measure performance (review speed, cache hit rates)
Calculate ROI (cost vs time saved)
Analyze findings by category and severity
Monitor providers for reliability and cost-effectiveness

Quick Start

Enable Analytics

Analytics are enabled by default. To configure:

# .env or GitHub Actions secrets
ANALYTICS_ENABLED=true
ANALYTICS_MAX_REVIEWS=1000  # Keep last 1000 reviews

Generate Dashboard

CLI Mode

# Generate HTML dashboard
mpr analytics generate

# Generate CSV export
mpr analytics generate --format csv

# Generate JSON export
mpr analytics generate --format json --days 7

# View summary in terminal
mpr analytics summary
mpr analytics summary --days 30

Docker Mode

# Enter container
docker exec -it mpr-review sh

# Generate dashboard
node dist/cli/index.js analytics generate

# View in browser
# Copy reports/analytics-dashboard.html to your machine
# or mount reports directory

GitHub Actions

- name: Generate Analytics Dashboard
  run: |
    npx mpr analytics generate

- name: Upload Dashboard
  uses: actions/upload-artifact@v3
  with:
    name: analytics-dashboard
    path: reports/analytics-dashboard.html

Dashboard Features

Summary Cards

The dashboard shows key metrics at a glance:

Total Reviews: Number of reviews completed
Total Cost: Cumulative cost across all reviews
Average Cost: Cost per review
Total Findings: Issues discovered across all reviews
Cache Hit Rate: Percentage of reviews using cache
ROI: Return on investment (time saved / cost)

Cost Trends Chart

Line chart showing:

Daily cost over time
Number of reviews per day
Cost per review trend

Use cases:

Identify cost spikes
Track cost optimization efforts
Budget forecasting

Performance Trends Chart

Track review performance:

Average review duration over time
Speed improvements from caching and incremental reviews

Use cases:

Measure optimization impact
Identify performance regressions
Developer experience improvements

Findings Distribution

Pie chart showing:

Findings by severity (Critical, Major, Minor)
Findings by category (Security, Performance, Style, etc.)

Use cases:

Focus engineering effort on common issues
Track improvement in code quality over time
Identify training opportunities

Provider Performance Table

Compare providers by:

Number of reviews
Success rate
Average cost
Average duration

Use cases:

Optimize provider selection
Identify unreliable providers
Cost/performance tradeoffs

ROI Analysis

Automatic ROI calculation:

Time Saved = Reviews × 30 minutes (avg manual review time)
Cost = Sum of all review costs
ROI = Time Saved / Cost

Example: 100 reviews × 30 min = 50 hours saved
         Cost: $2.50
         ROI: 50 hours / $2.50 = 20x return

CLI Commands

Summary Command

Show quick statistics in terminal:

# Last 30 days (default)
mpr analytics summary

# Custom time range
mpr analytics summary --days 7
mpr analytics summary --days 90

# Output:
# === Analytics Summary ===
#
# Total Reviews: 150
# Total Cost: $3.45
# Average Cost per Review: $0.0230
# Total Findings: 1,247
# Cache Hit Rate: 68.0%
#
# ROI:
#   Total Cost: $3.45
#   Estimated Time Saved: 75.0 hours
#   ROI: 2,174x
#
# Top Providers:
#   1. gemini-2.0-flash-exp: 85 reviews, 98.8% success
#   2. devstral-2512: 82 reviews, 96.3% success

Generate Command

Create analytics reports:

# HTML dashboard (default)
mpr analytics generate

# Specify output directory
mpr analytics generate --output ./custom-reports

# CSV export
mpr analytics generate --format csv

# JSON export
mpr analytics generate --format json

# Custom time range
mpr analytics generate --days 7 --format csv

Output files:

analytics-dashboard.html - Interactive HTML dashboard
analytics-export.csv - Spreadsheet-compatible data
analytics-metrics.json - Raw metrics for further processing

Data Storage

Location

Analytics data is stored in GitHub Actions cache:

Cache Key: analytics-metrics-data
Location: .cache/analytics/
Size: ~50KB per 1000 reviews

Data Retention

# Maximum reviews stored (prevents unbounded growth)
ANALYTICS_MAX_REVIEWS=1000  # Default

# Older reviews are automatically pruned
# Only the most recent N reviews are kept

Data Structure

interface ReviewMetric {
  timestamp: number;          // Unix timestamp
  prNumber: number;           // PR number
  filesReviewed: number;      // Files analyzed
  findingsCount: number;      // Total findings
  costUsd: number;            // Review cost in USD
  durationSeconds: number;    // Review duration
  providersUsed: number;      // Number of providers
  cacheHit: boolean;          // Whether cache was used
}

Export Data

# Export to CSV for analysis in Excel/Sheets
mpr analytics generate --format csv

# Export to JSON for custom processing
mpr analytics generate --format json

# Process JSON with jq
cat reports/analytics-metrics.json | jq '.[] | select(.costUsd > 0.05)'

Integration Examples

GitHub Actions Workflow

name: Weekly Analytics Report

on:
  schedule:
    - cron: '0 9 * * 1'  # Monday 9am
  workflow_dispatch:

jobs:
  analytics:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Generate Analytics
        run: |
          npx mpr analytics generate --days 7

      - name: Upload Dashboard
        uses: actions/upload-artifact@v3
        with:
          name: weekly-analytics
          path: reports/analytics-dashboard.html

      - name: Post Summary
        run: |
          npx mpr analytics summary --days 7 > analytics.txt
          gh issue create \
            --title "Weekly Analytics Report" \
            --body "$(cat analytics.txt)" \
            --label "analytics"
        env:
          GH_TOKEN: ${{ github.token }}

Slack Notifications

- name: Send to Slack
  run: |
    SUMMARY=$(npx mpr analytics summary --days 7)
    curl -X POST ${{ secrets.SLACK_WEBHOOK_URL }} \
      -H 'Content-Type: application/json' \
      -d "{\"text\":\"Weekly Code Review Analytics:\n\`\`\`$SUMMARY\`\`\`\"}"

Email Reports

- name: Email Report
  uses: dawidd6/action-send-mail@v3
  with:
    server_address: smtp.gmail.com
    server_port: 465
    username: ${{ secrets.MAIL_USERNAME }}
    password: ${{ secrets.MAIL_PASSWORD }}
    subject: Weekly Analytics Report
    to: team@company.com
    from: GitHub Actions
    body: file://analytics.txt
    attachments: reports/analytics-dashboard.html

Custom Processing

# Get cost by day
cat analytics-metrics.json | jq '
  group_by(.timestamp / 86400000 | floor) |
  map({
    date: .[0].timestamp,
    reviews: length,
    cost: map(.costUsd) | add
  })
'

# Find expensive reviews
cat analytics-metrics.json | jq '
  map(select(.costUsd > 0.05)) |
  sort_by(-.costUsd)
'

# Calculate monthly spend
cat analytics-metrics.json | jq '
  map(.costUsd) | add
'

Cost Optimization

Identify Cost Drivers

# Generate detailed report
mpr analytics generate --format csv

# Open in Excel/Sheets
# Pivot by: Provider, Date, PR Number
# Sum by: Cost

Optimize Provider Selection

Review provider performance:
- Check success rates
- Compare costs per provider
- Identify slow providers
Adjust provider list:

# Use only top-performing free providers
REVIEW_PROVIDERS=openrouter/google/gemini-2.0-flash-exp:free,openrouter/mistralai/devstral-2512:free

# Reduce provider count
PROVIDER_LIMIT=3

Enable cost controls:

# Set budget limit
BUDGET_MAX_USD=1.0

# Use more free providers
# See available free providers:
# https://openrouter.ai/models?order=newest&supported_parameters=tools&max_price=0

Leverage Caching

# Enable all caching features
ENABLE_CACHING=true
INCREMENTAL_ENABLED=true
GRAPH_ENABLED=true
GRAPH_CACHE_ENABLED=true

# Check cache hit rate in dashboard
# Target: >60% cache hit rate

Performance Monitoring

Track Review Speed

# View performance trends in dashboard
mpr analytics generate

# Check average duration
mpr analytics summary | grep "Duration"

Benchmark Against Targets

Metric	Target	Actual (Your Data)
Review Duration	<60s	Check dashboard
Cache Hit Rate	>60%	Check dashboard
Cost per Review	<$0.05	Check dashboard
Findings per Review	8-12	Check dashboard

Alerts

Set up alerts for anomalies:

# Check if cost is above threshold
COST=$(mpr analytics summary --days 1 | grep "Total Cost" | awk '{print $3}' | tr -d '$')
if (( $(echo "$COST > 1.0" | bc -l) )); then
  echo "⚠️ Daily cost exceeded $1.00: $$COST"
  # Send alert
fi

Privacy & Security

What Data is Collected?

Analytics collect only metadata:

✅ Review timestamp
✅ PR number
✅ File count
✅ Finding count
✅ Cost and duration
✅ Provider names
✅ Cache hit status

Not collected:

❌ Code content
❌ Finding details
❌ PR descriptions
❌ User names (except in PR number context)
❌ Repository names

Data Storage

Stored in GitHub Actions cache (if using Actions)
Stored in Docker volume (if self-hosted)
Never sent to external analytics services
Fully under your control

Disable Analytics

# Disable completely
ANALYTICS_ENABLED=false

# Or clear data
rm -rf .cache/analytics/

Troubleshooting

No Data in Dashboard

# Check if analytics is enabled
echo $ANALYTICS_ENABLED

# Check cache
ls -la .cache/analytics/

# Verify reviews have run
mpr analytics summary

Dashboard Not Generating

# Check for errors
LOG_LEVEL=debug mpr analytics generate

# Ensure reports directory exists
mkdir -p reports

# Check permissions
chmod 755 reports

Inaccurate Costs

Cost data comes from OpenRouter API
Free providers show $0.00 cost
Custom providers may not report costs
Check provider pricing: https://openrouter.ai/models

Best Practices

Generate reports weekly - Track trends over time
Set budget alerts - Prevent cost overruns
Monitor cache hit rates - Optimize for performance
Review provider performance - Remove underperforming providers
Export data regularly - Backup for long-term analysis
Share with team - Make data-driven decisions
Celebrate wins - Show ROI to stakeholders

Support

For analytics questions:

Example dashboard: reports/analytics-dashboard.html
GitHub Issues: https://github.com/777genius/review-router/issues
Analytics code: src/analytics/

License

MIT License - See LICENSE file for details

FilesExpand file tree

analytics.md

Latest commit

History

analytics.md

File metadata and controls

Analytics & Reporting Guide

Overview

Quick Start

Enable Analytics

Generate Dashboard

CLI Mode

Docker Mode

GitHub Actions

Dashboard Features

Summary Cards

Cost Trends Chart

Performance Trends Chart

Findings Distribution

Provider Performance Table

ROI Analysis

CLI Commands

Summary Command

Generate Command

Data Storage

Location

Data Retention

Data Structure

Export Data

Integration Examples

GitHub Actions Workflow

Slack Notifications

Email Reports

Custom Processing

Cost Optimization

Identify Cost Drivers

Optimize Provider Selection

Leverage Caching

Performance Monitoring

Track Review Speed

Benchmark Against Targets

Alerts

Privacy & Security

What Data is Collected?

Data Storage

Disable Analytics

Troubleshooting

No Data in Dashboard

Dashboard Not Generating

Inaccurate Costs

Best Practices

Support

License