Routine Stats

View metrics and analytics for routine executions. Track success rates, common issues, time trends, and get optimization suggestions.

Usage

/routine-stats                   # Overall summary
/routine-stats --merge           # Merge routine stats only
/routine-stats --detailed        # Detailed breakdown
/routine-stats --recent          # Last 10 executions
/routine-stats --trends          # Time-based trends
/routine-stats --checks          # Per-check statistics
/routine-stats --recommendations # Optimization suggestions

Arguments

Parse $ARGUMENTS for:

• --merge or -m - Show merge routine stats
• --detailed or -d - Detailed breakdown with charts
• --recent <n> or -r <n> - Show last N executions (default: 10)
• --trends or -t - Show trends over time
• --checks or -c - Per-check statistics
• --recommendations - Get optimization suggestions
• --export - Export stats as JSON/CSV
• --reset - Reset all statistics (asks for confirmation)
• --since <date> - Stats since specific date (e.g., --since 2026-01-01)
• --level <level> - Filter by routine level

---

Statistics Overview

Summary View (Default)

/routine-stats

Output:

# Routine Statistics

**Period**: Last 30 days
**Total Executions**: 47

## Merge Routine

**Executions**: 47
**Success Rate**: 93.6% (44/47)
**Avg Duration**: 4m 23s

### By Level
| Level | Executions | Success | Avg Time |
|-------|-----------|---------|----------|
| Light | 12 | 100% | 1m 45s |
| Medium | 28 | 92.9% | 4m 12s |
| Thorough | 7 | 85.7% | 9m 48s |

### Most Common Issues
1. GitHub Actions timeout (8 occurrences)
2. Gemini HIGH severity items requiring review (12 occurrences)
3. Security vulnerabilities found (5 occurrences)

### Trends
- ✅ Success rate improving (90% → 94% this week)
- ⚠️ Average time increasing (3m 50s → 4m 23s)
- ✅ Fewer security issues (8 → 5 this week)

---

Detailed Statistics

Recent Executions

/routine-stats --recent

Output:

# Recent Routine Executions (Last 10)

| Date | PR | Level | Duration | Status | Issues |
|------|-----|-------|----------|--------|--------|
| 01/30 20:05 | #456 | Medium | 4m 12s | ✅ Merged | 0 |
| 01/30 18:23 | #454 | Thorough | 12m 8s | ✅ Merged | 2 MEDIUM |
| 01/30 15:47 | #452 | Medium | 3m 54s | ✅ Merged | 1 HIGH |
| 01/30 14:12 | #450 | Light | 1m 32s | ✅ Merged | 0 |
| 01/30 11:05 | #448 | Medium | TIMEOUT | ❌ Failed | Gemini timeout |
| 01/29 19:34 | #446 | Thorough | 10m 23s | ✅ Merged | 3 HIGH |
| 01/29 17:12 | #444 | Medium | 4m 8s | ✅ Merged | 0 |
| 01/29 15:28 | #442 | Light | 1m 48s | ✅ Merged | 0 |
| 01/29 13:45 | #440 | Medium | 5m 2s | ⚠️ Blocked | CRITICAL security |
| 01/29 10:15 | #438 | Medium | 4m 15s | ✅ Merged | 1 MEDIUM |

**Click any PR** to view detailed execution log

Per-Check Statistics

/routine-stats --checks

Output:

# Check Performance

| Check | Runs | Pass Rate | Avg Time | Failures | Auto-Fixed |
|-------|------|-----------|----------|----------|------------|
| Build | 47 | 100% | 8.2s | 0 | - |
| Tests | 47 | 97.9% | 12.3s | 1 | 0 |
| TypeScript | 35 | 94.3% | 9.8s | 2 | 0 |
| Lint | 35 | 91.4% | 4.2s | 3 | 3 |
| Secrets Scan | 47 | 100% | 1.2s | 0 | - |
| Gemini Review | 35 | 88.6% | 45.2s | 4 (timeout) | - |
| Security Scan | 42 | 85.7% | 18.4s | 6 | 4 |
| GitHub Actions | 35 | 80.0% | 2m 14s | 7 | 0 |
| Bundle Size | 7 | 85.7% | 32.1s | 1 | 0 |
| Performance | 7 | 71.4% | 48.3s | 2 | 0 |

**Issues:**
- Gemini Review: 4 timeouts (increase timeout?)
- GitHub Actions: 7 failures (investigate workflows)
- Performance: 2 failures (Lighthouse < 90)

---

Trends & Analytics

Time Trends

/routine-stats --trends

Output:

# Routine Trends

## Success Rate Over Time

Week 1: ████████░░ 80% (8/10) Week 2: █████████░ 90% (9/10) Week 3: ██████████ 100% (10/10) Week 4: █████████░ 92% (11/12)

**Trend**: ✅ Improving (+15% month-over-month)

## Average Duration

Week 1: ████░░░░░░ 4m 45s Week 2: ████░░░░░░ 4m 12s Week 3: ███░░░░░░░ 3m 58s Week 4: ████░░░░░░ 4m 23s

**Trend**: ⚠️ Slightly increasing this week (+6%)

## Issues Found

CRITICAL: ▂▁▁▁▁▁▁▁ (declining - good!) HIGH: ▃▄▃▂▃▂▁▂ (stable) MEDIUM: ▅▆▄▅▃▄▅▆ (slightly increasing) LOW: ▇▆▇▅▆▇▆▅ (stable)

## Auto-Fix Success

Lint: ████████████ 100% (12/12) Security: ██████░░░░░░ 66% (4/6) Overall: ████████░░░░ 88% (16/18)

---

Optimization Recommendations

/routine-stats --recommendations

Output:

# Optimization Recommendations

Based on 47 routine executions in the last 30 days:

## 1. Increase Gemini Timeout
**Current**: 180s
**Recommended**: 240s
**Reason**: 4 timeouts in last 30 days (11% of Gemini reviews)
**Impact**: Reduce failures by ~8%

Apply:
/routine-config --set checks.gemini_review.timeout_seconds 240

---

## 2. Consider Skipping Lint for Light Level
**Current**: Lint enabled for all levels
**Recommended**: Disable for light level
**Reason**: 100% pass rate, adds 4.2s, rarely finds issues in docs/tests
**Impact**: Reduce light routine time by ~20%

Apply:
/routine-config --set checks.lint.levels '["medium","thorough"]'

---

## 3. Enable Auto-Fix for Security
**Current**: auto_fix_npm_audit = false
**Recommended**: auto_fix_npm_audit = true
**Reason**: All 4 npm audit issues were auto-fixable
**Impact**: Reduce manual intervention by ~10%

Apply:
/routine-config --set checks.security_scan.auto_fix_npm_audit true

---

## 4. Use Preset for Dependency Updates
**Pattern Detected**: 8 PRs with titles containing "bump", "update deps"
**Recommendation**: Create/use deps-update preset
**Impact**: Optimized checks for dependency PRs, save ~2m per execution

Use:
/routine-merge --preset deps-update

---

## 5. Investigate GitHub Actions Failures
**Issue**: 7 failures in 35 runs (20% failure rate)
**Common failures**:
  - E2E tests (4 failures)
  - Deploy preview (2 failures)
  - Linting (1 failure)
**Recommendation**: Review workflow stability
**Impact**: Improve success rate by ~15%

---

## 6. Consider Default Level Adjustment
**Current Default**: medium
**Analysis**:
  - Light: 12 runs, 100% success, fast (1m 45s)
  - Medium: 28 runs, 93% success, moderate (4m 12s)
  - Thorough: 7 runs, 86% success, slow (9m 48s)

**Recommendation**:
  - 40% of medium runs had zero issues → could use light
  - Keep medium as default for safety
  - Consider auto-suggesting level based on PR

---

## Summary
- Implementing all recommendations could:
  - ✅ Reduce average execution time by ~18%
  - ✅ Improve success rate by ~12%
  - ✅ Reduce manual interventions by ~15%

---

Export Statistics

JSON Export

/routine-stats --export

Creates: ~/.claude/routines/stats-export-{timestamp}.json

{
  "exported_at": "2026-01-30T20:30:00Z",
  "period": {
    "start": "2026-01-01T00:00:00Z",
    "end": "2026-01-30T23:59:59Z"
  },
  "summary": {
    "total_executions": 47,
    "success_rate": 0.936,
    "avg_duration_seconds": 263,
    "by_level": {
      "light": {
        "count": 12,
        "success_rate": 1.0,
        "avg_duration_seconds": 105
      },
      "medium": {
        "count": 28,
        "success_rate": 0.929,
        "avg_duration_seconds": 252
      },
      "thorough": {
        "count": 7,
        "success_rate": 0.857,
        "avg_duration_seconds": 588
      }
    }
  },
  "checks": [...],
  "issues": [...],
  "trends": [...]
}

CSV Export

/routine-stats --export --format csv

Creates: ~/.claude/routines/stats-export-{timestamp}.csv

Useful for:

• Importing to spreadsheets
• Creating custom charts
• Sharing with team
• Compliance/audit trails

---

Filter Statistics

By Date Range

/routine-stats --since 2026-01-15

Shows stats from Jan 15 onwards.

By Level

/routine-stats --level thorough

Shows only thorough routine stats.

By Status

/routine-stats --status failed

Shows only failed routines (for debugging).

---

Reset Statistics

/routine-stats --reset

Warning:

This will permanently delete all routine statistics.

Stats include:
- 47 routine executions
- Per-check metrics
- Trend data
- Optimization insights

Execution state files will be preserved.

Create backup first? [Yes/No/Cancel]

If Yes:

1. Creates backup: ~/.claude/routines/stats-backup-{timestamp}.json
2. Deletes statistics
3. Preserves state files

---

State File Details

View detailed execution log:

/routine-stats --pr 456

Output:

# Routine Execution: PR #456

**Status**: ✅ Merged
**Level**: Medium
**Branch**: feature/auth → main
**Started**: 2026-01-30 20:00:15
**Completed**: 2026-01-30 20:04:27
**Duration**: 4m 12s

## Timeline

20:00:15 - Started medium routine
20:00:23 - ✅ Build passed (8.2s)
20:00:35 - ✅ Tests passed (12.1s, 94% coverage)
20:00:37 - 🔄 Starting Gemini review...
20:01:22 - ✅ Gemini review completed (45.3s)
             - 2 CRITICAL items auto-implemented
             - 1 HIGH item user confirmed
             - 3 MEDIUM items deferred
20:01:35 - ✅ Security scan passed (13.2s)
20:01:45 - ✅ TypeScript check passed (9.8s)
20:01:50 - ✅ Lint passed (4.2s)
20:02:15 - 🔄 Waiting for GitHub Actions...
20:04:20 - ✅ GitHub Actions passed (2m 5s)
20:04:25 - User confirmed merge
20:04:27 - ✅ PR merged (squash: abc123)

## Changes Made
1. fix: sanitize user input (triggered by Gemini CRITICAL)
   Commit: abc123

2. fix: update bcrypt to v5.1.1 (triggered by Security HIGH)
   Commit: def456

## Deferred Items
- [MEDIUM] Consider adding rate limiting → Tracked in #457
- [MEDIUM] Improve error messages
- [LOW] Add more test cases

## Metrics
- Total checks: 8
- Checks passed: 8
- Auto-fixes applied: 2
- User interventions: 1
- Re-runs required: 2 (after implementing changes)

---

Learning Insights

The statistics system provides insights like:

Pattern Recognition

Detected Pattern: PRs from user @alice rarely have issues
Suggestion: Consider light level for @alice's PRs

Check Optimization

Insight: TypeScript check never fails for test-only PRs
Suggestion: Skip TypeScript for PRs touching only *.test.ts files

Timing Optimization

Insight: Gemini review averages 45s, timeout is 180s
Suggestion: Reduce timeout to 90s for faster failures

---

Examples

Example 1: Weekly Review

User: /routine-stats --since 2026-01-24

Claude: [Shows stats for last week]

**Week Summary**:
- 12 executions
- 92% success rate
- 4m 15s average

**Improvements**:
- Success rate up 5% from previous week
- Average time down 8s

**Issues**:
- 1 Gemini timeout (increased timeout recommended)

Example 2: Check Specific Issue

User: /routine-stats --checks

Claude: [Shows per-check stats]

User: Why is GitHub Actions failing 20%?

Claude: Analyzing GitHub Actions failures...

7 failures in 35 runs:
1. E2E tests failing on webkit (4×)
2. Deploy preview timeout (2×)
3. Lint job OOM (1×)

Recommendations:
- Fix webkit E2E flakiness
- Increase deploy timeout
- Increase lint job memory

Example 3: Optimization

User: /routine-stats --recommendations

Claude: [Shows 6 recommendations]

User: Apply recommendations 1 and 3

Claude: Applying:
1. Increase Gemini timeout: 180s → 240s
3. Enable auto-fix for security

✅ Configuration updated
✅ Changes will apply to next routine execution

Would you like to test with a dry run?
/routine-merge --dry-run

---

Related Commands

• /routine-merge - Run merge routine
• /routine-config - Configure routine settings

---

Notes

• Statistics are updated after each routine execution
• Trends require minimum 5 executions
• Recommendations require minimum 10 executions
• Export includes all raw data for custom analysis
• Reset preserves execution state files (only deletes aggregated stats)