How I Took ~1,000 Pages From Hundreds of Rich Results Warnings to Zero

Description and use

A PHP script automated the detection of Rich Results warnings across ~1,000 pages by validating required properties, @id references, and JSON‑LD syntax, and was integrated into CI to prevent regressions.

Problem

Hundreds of Rich Results warnings appeared on ~1,000 pages of structured data because JSON‑LD references pointed to nodes that were never defined.

Outcome

The auditor reduced Rich Results warnings from hundreds to zero across ~1,000 pages and provides continuous protection against future regressions.

Workflow

• 1. Extract all <script type="application/ld+json"> blocks from each rendered HTML file.
• 2. Run three checks: required properties, @id resolution, and JSON parse errors.
• 3. First pass indexes all nodes with @id to build a definition map.
• 4. Second pass walks each node recursively, validating internal @id references against the map.
• 5. Emit errors for missing properties, unresolved references, or parse failures.
• 6. Return exit code 0 if clean, 1 if errors, 2 if misuse.
• 7. Integrate the script into CI so any new warning causes a build failure.

Evidence

Google flagged hundreds of Rich Results warnings across them.

Source

jsonld-schema-audit is one PHP file with no dependencies.

Source

The exit code ties it together: 0 when clean, 1 when there are errors, 2 for bad usage.

Source