Skip to content

performance-audit.md

Latest commit

 

History

History
175 lines (122 loc) · 4.11 KB

performance-audit.md

File metadata and controls

175 lines (122 loc) · 4.11 KB

Performance Audit Recipe

WPCC → WP Performance Timer Pipeline

Note: WP Performance Timer will be renamed to Hypercart WP Performance Timer in a future release.

This recipe combines static analysis (WPCC) with runtime profiling (WP Performance Timer) to identify and confirm performance bottlenecks.

When to Use

  • Pre-deploy performance audit
  • Investigating slow page loads
  • Confirming WPCC performance warnings
  • Before/after optimization comparison

Prerequisites

  1. WPCC available via wpcc command (part of AI-DDTK)
  2. WP Performance Timer plugin installed and activated in WordPress
  3. Access to WordPress debug.log or NeoLog session data

Workflow

Phase 1: Static Analysis (WPCC)

Scan the target code for potential performance issues:

wpcc --paths /path/to/plugin --format json

What to look for:

  • n-plus-1-pattern - Queries inside loops
  • unbounded-query - Missing LIMIT clauses
  • expensive-operation - Heavy computations

Output: JSON file in tools/wp-code-check/dist/logs/[TIMESTAMP].json

Phase 2: AI Triage

Review findings and prioritize:

Ask AI: "Triage this scan"

Focus on:

  • High-severity performance warnings
  • Code that runs on every page load
  • Database queries in loops

Phase 3: Runtime Profiling

For each confirmed issue, add performance timers:

// Before the suspected slow code
if (function_exists('perf_timer_start')) {
    $timer = perf_timer_start('suspect-operation', [
        'file' => __FILE__,
        'line' => __LINE__,
        'wpcc_finding' => 'n-plus-1-pattern'
    ]);
}

// The code WPCC flagged
foreach ($items as $item) {
    $result = get_post_meta($item->ID, 'some_meta', true);
}

// After the suspected slow code
if (isset($timer) && function_exists('perf_timer_stop')) {
    perf_timer_stop($timer);
}

Phase 4: Trigger and Measure

  1. Enable verbose logging (optional):

    // In wp-config.php
define('PERF_LOG_ALL', true);

  2. Trigger the operation:

    • Load the page
    • Trigger the AJAX call
    • Run the cron job

  3. Collect metrics from:

    • wp-content/debug.log
    • NeoLog session files
    • Admin → Tools → Performance Logs

Phase 5: Report

Document findings in this format:

## Performance Audit Report

### Summary
| Metric | Value |
|--------|-------|
| Target | checkout-page |
| Date | 2026-02-02 |
| WPCC Findings | 3 potential issues |
| Confirmed Bottlenecks | 1 |

### Confirmed Issues

#### Issue 1: Loop at line 234
- **WPCC Finding**: n-plus-1-pattern
- **Runtime Impact**: 1,847ms, 156 queries
- **Recommendation**: Batch query with single get_posts() call

### False Positives

#### WPCC Finding: unbounded-query (line 89)
- **Runtime Measurement**: 12ms, 1 query
- **Reason**: Result set is naturally small (max 5 items)
- **Action**: No change needed

Quick Commands

# Phase 1: Scan
wpcc --paths <path> --format json

# Check if Performance Timer is enabled
wp eval "var_dump(defined('PERF_TIMING_ENABLED') && PERF_TIMING_ENABLED);"

# View recent performance logs
tail -100 /path/to/wordpress/wp-content/debug.log | grep PERF

Tips

  1. Profile the hot path first - Focus on code that runs on every request
  2. Use hierarchy - Wrap large operations, then add child timers for sub-operations
  3. Compare before/after - Always measure before optimizing, then measure again
  4. Watch query counts - High query counts in loops are the #1 WordPress performance killer
  5. Check memory too - Memory leaks cause issues on long-running operations

Related

AI Agent Instructions

When user asks for a performance audit:

  1. Run WPCC scan on the target path
  2. Triage findings for performance-related issues
  3. Guide user to add timers around flagged code
  4. Help interpret the runtime metrics
  5. Provide optimization recommendations based on confirmed bottlenecks