agentleFS
iwritec0deiwritec0de/wp-dev/skills/wordpress-performance/SKILL.md

wordpress-performance

Skill1 starsChanged 7 months agoMITView on GitHub

This skill should be used when the user asks to "optimize WordPress queries", "audit plugin performance", "reduce page load time", "check autoloaded options", "profile WordPress hooks", "fix slow queries", "add caching", or mentions "WordPress performance", "query optimization", "SAVEQUERIES", "object cache", "transients", "autoload", "Query Monitor", "slow query", "N+1", "database optimization", "page speed", "TTFB", "hook profiling". Provides WordPress performance expertise covering query optimization, caching strategies, hook profiling, autoload management, and live site diagnostics.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If you frequently filter by a specific meta key, consider adding an index:

    line 227

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

WordPress Performance

Performance optimization patterns for WordPress plugins, themes, and live sites.

Critical Rules

  1. Never query without limits — posts_per_page => -1 and unbounded $wpdb queries are the most common cause of memory exhaustion on production sites.
  2. Cache expensive queries — any query that runs on every page load and doesn't change per-request must be cached with transients or the object cache.
  3. Use no_found_rows — set 'no_found_rows' => true on any WP_Query that doesn't need pagination; this skips SQL_CALC_FOUND_ROWS which is expensive on large tables.
  4. Avoid meta queries on unindexed keys — meta_query on wp_postmeta performs full table scans unless you add custom indexes. Consider a custom table for heavily queried data.
  5. Defer heavy work — operations on init run on every request. Use admin_init for admin-only work, rest_api_init for REST-only, and wp_loaded or later hooks when possible.
  6. Prime caches, don't loop-query — use update_post_meta_cache(), update_post_caches(), or _prime_post_caches() to batch-load metadata instead of calling get_post_meta() in a loop.

Query Performance Patterns

Efficient WP_Query

// Good — limited, cache-friendly, no unnecessary data:
$query = new WP_Query( [
    'post_type'              => 'product',
    'posts_per_page'         => 20,
    'no_found_rows'          => true,   // Skip pagination count query.
    'update_post_meta_cache' => false,  // Skip if not reading meta.
    'update_post_term_cache' => false,  // Skip if not reading terms.
    'fields'                 => 'ids',  // Return only IDs when full objects aren't needed.
] );

// Bad — unbounded, forces full table scan:
$query = new WP_Query( [
    'post_type'      => 'product',
    'posts_per_page' => -1,  // Never do this.
] );

N+1 Query Prevention

// Bad — N+1 pattern (1 query per post in the loop):
foreach ( $posts as $post ) {
    $price = get_post_meta( $post->ID, '_price', true );  // Query per iteration.
}

// Good — prime the cache first, then loop reads from cache:
$post_ids = wp_list_pluck( $posts, 'ID' );
update_meta_cache( 'post', $post_ids );  // Single query loads all meta.

foreach ( $posts as $post ) {
    $price = get_post_meta( $post->ID, '_price', true );  // Reads from cache.
}

Batch Operations

// Bad — individual inserts in a loop:
foreach ( $items as $item ) {
    $wpdb->insert( $table, $item );  // N queries.
}

// Good — single bulk insert:
$values = [];
foreach ( $items as $item ) {
    $values[] = $wpdb->prepare( '(%d, %s, %f)', $item['user_id'], $item['status'], $item['total'] );
}
if ( $values ) {
    $wpdb->query( "INSERT INTO {$table} (user_id, status, total) VALUES " . implode( ', ', $values ) );
}

Caching Strategies

Transient Caching Pattern

function myplugin_get_featured_products(): array {
    $cache_key = 'myplugin_featured_products';
    $data      = get_transient( $cache_key );

    if ( false !== $data ) {
        return $data;
    }

    $query = new WP_Query( [
        'post_type'      => 'product',
        'posts_per_page' => 12,
        'meta_key'       => '_featured',
        'meta_value'     => 'yes',
        'no_found_rows'  => true,
        'fields'         => 'ids',
    ] );

    $data = $query->posts;
    set_transient( $cache_key, $data, HOUR_IN_SECONDS );

    return $data;
}

// Invalidate when products change:
add_action( 'save_post_product', function (): void {
    delete_transient( 'myplugin_featured_products' );
} );

Object Cache for Per-Request Deduplication

function myplugin_get_settings(): array {
    $cached = wp_cache_get( 'settings', 'myplugin' );
    if ( false !== $cached ) {
        return $cached;
    }

    $settings = get_option( 'myplugin_settings', [] );
    wp_cache_set( 'settings', $settings, 'myplugin' );

    return $settings;
}

When to Use Each Layer

ScenarioStrategy
Same data fetched multiple times per requestwp_cache_* (object cache)
Expensive query, result valid for minutes/hoursset_transient()
External API responseset_transient() with timeout matching API rate limits
Data that changes on saveTransient + delete_transient() on save_post
User-specific dataTransient with user ID in key, or get_user_meta()
Full-page outputConsider wp_cache_* with a persistent backend (Redis/Memcached)

Hook Performance

Defer Initialization

// Bad — runs on every request including REST, AJAX, cron:
add_action( 'init', 'myplugin_heavy_init' );

// Good — scope to context:
add_action( 'admin_init', 'myplugin_admin_setup' );      // Admin only.
add_action( 'rest_api_init', 'myplugin_register_routes' ); // REST only.
add_action( 'template_redirect', 'myplugin_frontend' );    // Frontend only.

// Good — conditional loading:
add_action( 'init', function (): void {
    if ( ! is_admin() && ! wp_doing_ajax() && ! wp_doing_cron() ) {
        // Frontend-only logic.
    }
} );

Lazy Loading

// Bad — always loads all classes:
require_once __DIR__ . '/includes/class-admin.php';
require_once __DIR__ . '/includes/class-reports.php';
require_once __DIR__ . '/includes/class-import.php';

// Good — load only when needed:
add_action( 'admin_menu', function (): void {
    require_once __DIR__ . '/includes/class-admin.php';
    new Myplugin_Admin();
} );

// Best — PSR-4 autoloader via Composer:
require_once __DIR__ . '/vendor/autoload.php';

Autoload Optimization

Options with autoload = yes are loaded into memory on every page load via a single SELECT from wp_options. Large or rarely-used options bloat this query.

// Bad — large data autoloaded:
add_option( 'myplugin_log_history', $huge_array );  // Defaults to autoload = yes.

// Good — disable autoload for large or infrequent data:
add_option( 'myplugin_log_history', $huge_array, '', false );

// Or with update_option (autoload param is 4th arg since WP 4.2):
update_option( 'myplugin_log_history', $huge_array, false );

Rule of thumb: autoload only small, frequently-accessed options (settings, feature flags). Disable autoload for logs, large serialized arrays, and data accessed only on specific pages.

Database Indexing

// Add indexes to custom tables on activation:
register_activation_hook( __FILE__, function (): void {
    global $wpdb;
    $table = $wpdb->prefix . 'myplugin_orders';

    // Check before adding to make activation idempotent.
    $index_exists = $wpdb->get_var( $wpdb->prepare(
        'SELECT COUNT(*) FROM INFORMATION_SCHEMA.STATISTICS WHERE TABLE_SCHEMA = %s AND TABLE_NAME = %s AND INDEX_NAME = %s',
        DB_NAME, $table, 'status_created_idx'
    ) );

    if ( ! $index_exists ) {
        $wpdb->query( "ALTER TABLE {$table} ADD INDEX status_created_idx (status, created_at)" );
    }
} );

Indexing Post Meta for Queries

If you frequently filter by a specific meta key, consider adding an index:

-- Only add this if meta_query on this key is a known bottleneck:
ALTER TABLE wp_postmeta ADD INDEX meta_key_value_idx (meta_key(50), meta_value(50));

Warning: modifying core tables is risky and may conflict with updates. Prefer custom tables for heavily queried structured data.

Anti-Patterns

Anti-PatternImpactFix
posts_per_page => -1Loads unlimited posts into memorySet a reasonable limit or paginate
get_post_meta() in a loopN+1 queriesupdate_meta_cache() before loop
Heavy logic on initRuns on every requestUse context-specific hooks
get_option() on every function callRedundant queries (without persistent cache)Cache in a static variable or wp_cache_*
Large serialized arrays in autoloaded optionsBloats the autoload querySet autoload to false
$wpdb->query() inside foreachN inserts instead of 1Batch insert with concatenated VALUES
LIKE '%search%' on large tablesCannot use index (leading wildcard)Full-text search index or prefix-only LIKE
Missing no_found_rowsExtra COUNT query when pagination isn't neededSet true on non-paginated queries
wp_remote_get() on every page loadBlocks rendering on external serviceCache response with transient
Autoloading large optionsLoaded on every request even when unusedadd_option() with autoload false

For query analysis and EXPLAIN usage, see reference/query-optimization.md. For caching architecture and invalidation strategies, see reference/caching-strategies.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.