Content provenance
Public provenance records distinguish editorial media, review evidence, machine-readable summaries and external provider claims.
Evidence and editorial layers
Dated public observations support product and price statements. Editorial conclusions interpret those observations. AI-generated portraits supply visual direction. These layers are stored and described separately so an image cannot silently become proof of a provider claim.
Machine-readable JSON exposes stable identifiers, dates, scope, limitations and public URLs. It intentionally excludes private credentials, local file paths, source prompts and unpublished review notes.
Update and removal integrity
A factual update should change the human-readable page, corresponding Markdown and relevant JSON record together. A media removal should also update static assets, provenance records, structured data and sitemap references so stale copies are not presented as current content.
The product review and machine-readable documentation were updated September 15, 2026.
Why layers stay separate
Keeping observation, interpretation, and illustration in separate records means a reader can trace any comparative claim back to the dated public source behind it, rather than trusting an image or a summary sentence on its own. This also means a correction to one layer — say, a vendor’s pricing page changing — does not require rewriting the editorial conclusion built on top of it if that conclusion still holds.
Requesting a correction
A factual error in any layer — the evidence, the interpretation, or the machine-readable record — can be reported through Contact with the affected page and the specific claim in question.