Maintainer FAQ — Interpretation, reproducibility and limits

[mhdteto]

Are the demonstration results real client or AI-engine measurements?

No. Alex Rowan, Cedar Notes, publishers, page responses, reviews, engines and answers are invented teaching inputs. The reports are actual outputs of the linked code. The separate source catalogue contains real public entry points with dated access and verification notes; it is not part of the fictional evidence record.

Can I use the interactive explorer without Python?

Yes. The GitHub Pages explorer filters already published teaching data in your browser. It does not execute Python, upload inputs, crawl websites or query AI engines. The complete source example reproduces the reports locally.

Does “50% cited-answer rate” mean 50% visibility in a market?

No. In the fictional English/UK group, 2 of 4 supplied answers contain explicit citations. In the fictional French/France group, 1 of 2 does. Both fractions happen to equal 50%; neither is a market estimate, a product comparison, a causal result or an improvement claim. Keep engine, model, interface, prompt set, date, language and market in view.

Does an identity match prove the profile is true?

No. It means the supplied field agrees with the approved reference under the tool's documented normalization and alias rules. Check the underlying source and approval context. Descriptions require human review.

Does a reviewed evidence label or a URL prove a claim?

No. Support and relationship labels are supplied reviewer declarations. The mapper preserves contradictions, owned sources, unknown relationships, unreviewed material and missing evidence. It does not read the pages or decide truth.

Does a clean JSON-LD report establish Google eligibility or AI citations?

No. The kit checks an educational structural subset. Review full vocabulary rules, visible-page consistency, factual support and the applicable provider requirements separately.

What is the difference between an issue and a discussion?

Use issues for reproducible defects and scoped changes. Use Q&A for interpretation, Ideas for open proposals and Show and tell for documented examples. The roadmap groups accepted work and proposals; planned versions and timing are provisional.

Which versions and licenses apply?

The six tools have v1.0.0 releases with pinned source archives. The bundled workflow records exact source commits and Git blob hashes. Tool code is MIT; original docs, data examples and artwork use the included CC BY 4.0 notices. Review the license files in each source repository, and the underlying rights of any external material you collect.

Where should I start for an actual project?

Read the measurement protocol, define the entity and material claims, obtain an approved reference and preserve dated observations. The public tutorials explain the workflow; a private assessment is a separate scoped advisory engagement.