Sitewide structured-data governance
Cross-Page Schema Entity Conflict Auditor
Join JSON-LD from many pages into one explainable entity graph. Find shared IDs with conflicting values, duplicate Product emitters, broken references, unstable Organization IDs, repeated blocks, and canonical mismatches.
- 1Add data
- 2Map and review
- 3Configure
- 4Analyze
- 5Review and export
Add data
Choose files or paste an export
Accepts HTML files, a ZIP of supported files, crawl CSV, JSON, JSONL, and NDJSON. Limits: 100 MB expanded, 2,000 files, and 1,000,000 property assertions per analysis.
Map crawl fields
Confirm the CSV columns
Select the page URL and every column containing extracted JSON-LD. Optional provenance fields make conflicts easier to trace.
Parse review
Check data health before analysis
Invalid blocks stay visible as recoverable errors. Mislabeled scripts are ignored rather than executed.
Review invalid blocks and records
| Page or source | Block | Error |
|---|
Configure
Choose identity scope and conflict properties
Leave all types selected for a full governance audit. Narrow the selection when investigating one emitter or template.
Preparing analysis…
Review and export
Cross-page entity evidence
Results are deterministic and tied to the active type and property configuration.
Export active evidence
Exports use the active search, severity, issue type, and result view. Spreadsheet-leading formulas are escaped in CSV.
What this cross-page audit can and cannot prove
Compares normalized IDs and selected property assertions while preserving each page, block, source, and template.
Suggests strong-key candidates but does not merge anonymous entities or guess which plugin is at fault.
Separates parse validity and structural evidence from Schema.org or Google feature eligibility.
Supported inputs and deliberate limits
HTML extraction supports inert application/ld+json scripts containing one object, arrays, and @graph. Crawl exports can map URL, canonical, title, template, status, indexability, date, source signature, and multiple JSON-LD columns. JSON and NDJSON records can use url plus jsonld or blocks.
This version does not parse Microdata or RDFa, execute JavaScript, fetch live URLs, repair JSON5 syntax, inspect encrypted archives, or claim that a reported conflict is a search-engine violation.
Frequently asked questions
Reviewed Jul 2026 · Sources and limitations
Review details: 2026-07-13 · Marc LaClear · v1.0
Reference sources:
Known limits:
- This tool analyzes only the files and extracted fields you provide. It does not fetch live pages, execute JavaScript, inspect CMS configuration, or prove which component is responsible without supplied provenance.
- Findings describe conflicts and implementation evidence; they do not guarantee indexing, rankings, rich results, or Google feature eligibility.
- Anonymous entity matches remain candidates until a person confirms identity. Weak similarity is deliberately excluded.
- Encrypted ZIP, ZIP64, Microdata, RDFa, JSON5 comments, trailing commas, and automatic repair of invalid JSON-LD are not supported in this version.