| name | magento2-performance-audit |
| description | This skill should be used when the user asks to "audit performance", "check Core Web
Vitals", "run Lighthouse", "check server configuration", "verify Redis/Varnish setup",
"analyze database queries", "find N+1 query issues", "review indexer configuration", "check
cron health", "debug cache flush", asks "why does full_page cache keep flushing", wants to
"trace FPC invalidation", or reports "too many ajax requests", a "customer data section
reload storm", or a "crawler overloading server". Performs a comprehensive performance and
health audit for Magento 2 projects against Adobe Commerce Best Practices. DEPENDENT on
magento2-dev-core for code-level performance patterns.
|
| compatibility | claude, codex, opencode, copilot, dsh |
| depends | ["magento2-dev-core"] |
| metadata | {"audience":"developers","workflow":"magento"} |
Magento 2 Performance Audit
This skill performs a comprehensive audit of Magento 2 performance, infrastructure, and code-level patterns.
Govard-Native Audit Coverage
govard audit run executes PHPCS and PHPStan through Govard's pinned lint
toolchain image — it covers coding-standard and static-analysis findings only.
As of Govard v1.64.0 no performance audit check exists:
govard audit run --checks performance fails with "audit check ... is not
implemented". Govard v1.64.0 does add a native check
() that machine-captures the stock profiler CSV for one
URL — a quick complement to this skill's manual per-page audit, not a replacement: it ships
no query log, no cross-page matrix, and no threshold analysis. The profiler requires a project target (not standalone), an absolute http(s) URL (the request carries so stock Magento enables the CSV), and is guarded by a per-project lease; the CSV lands as with its SHA in — open it as spreadsheet to read per-timer costs. Manual per-page captures (7 pages) cost ~2.5-3 min on reference project (50k queries/page with call-stack, 16-26s each) — keep all 7, run with 300s timeout or background polling and trap restore, not by sampling fewer categories (that hides per-item N+1s). Keep running this checklist
yourself and treat as the shared lint gate. Never present a
lint-only pass as a performance verdict.