Fix schema-audit CLI checking no pages

[jasperf]

Version: 5.24.3

This release, 5.24.3, fixes two defects that made schema auditing report wrong results on real WordPress sites, one in the wp-ops schema-audit CLI script and one in the schema_audit MCP tool. The CLI script added paths that start with / to a site URL that already ended in a slash, so every request went to //path/. Sites answered with a 301, every page was skipped, and on imagewize.com the script checked 0 of 11 pages. The script has been rebuilt to match the 5.24.2 MCP fix: it now takes its page list from the site's sitemap, lists non-200 URLs separately, and fetches each page once. The MCP tool read only the top-level @type, so the Organization, WebSite and BreadcrumbList nodes that SEO plugins nest in an @graph array went uncounted. It now walks the whole JSON-LD tree.

Schema Audit CLI Rebuild:

• Fixed the double-slash URL bug that led the script to skip every page, homepage included, as "not found."
• Pages now come from the site's sitemap, trying wp-sitemap.xml, then sitemap_index.xml, then sitemap.xml. From a sitemap index it reads only the page sitemaps. The homepage is always checked first, and the common paths are used only when a site has no sitemap. The output says when that fallback is used.
• Added a --max-pages flag (default 25) to cap how many sitemap pages are checked. The report says when the cap cuts the list short.
• URLs that don't return 200 now appear in a separate "NOT CHECKED" section with their status code and redirect target, and are left out of the totals.
• Fixed the "PAGES NEEDING SCHEMA MARKUP" list, which was always empty: the grep -c ... || echo "0" fallback printed 0 twice and broke the numeric test. The list is now built during the audit.
• Each page is now fetched once instead of up to six times, and --pages accepts full URLs as well as paths.
• JSON-LD blocks that span several lines or carry extra attributes are now parsed, and Person is now a detected type. This parsing needs perl, which comes with macOS and standard Linux servers.

MCP schema_audit Tool:

• checkSchemaTypes in mcp-server/src/tools/schemaAudit.ts now collects every @type value anywhere in the JSON-LD tree. This catches the nodes that Yoast, Rank Math and The SEO Framework nest inside @graph.
• Fixed the text fallback for JSON-LD that doesn't parse, which never matched because it lowercased the schema but not the type it searched for. It now uses a regex match that also covers @type arrays.

Catalog and Documentation:

• Updated the wp-cli/seo/schema-audit entry in go/internal/catalog/catalog.json with the new description, the perl requirement, the new --max-pages argument, and the updated --pages wording.
• Rewrote the schema audit section of wp-cli/seo/README.md to cover sitemap discovery, the page cap, the non-200 handling and the full list of detected types. Added usage examples for --max-pages, --pages and --output.
• Added the 5.24.3 entry to CHANGELOG.md.

Files Changed:

• CHANGELOG.md (Modified)
• go/internal/catalog/catalog.json (Modified)
• mcp-server/src/tools/schemaAudit.ts (Modified)
• wp-cli/seo/README.md (Modified)
• wp-cli/seo/schema-audit.sh (Modified)