OrientLore Reference Release

The scope, methods, verification record and unresolved human review of OrientLore’s reference-release candidate.

Last updated

A reference candidate, not a certification.

Candidate · Human content review pending

This milestone gathers the English-language reference library, deterministic calendar tools, reading paths and evidence practices developed through the first 100-step programme. It is not a claim of a fully reviewed encyclopedia, institutional endorsement or comprehensive coverage of Chinese culture.

This record describes work prepared and checked locally. A repository commit, a build and a public deployment are different events; this page is not evidence that a version is currently live on a production server.

Use it to
Convert and understand dates, locate an animal year, explore objects and named tellings, prepare considerate questions, and follow original sources.
Do not use it to
Diagnose, invest, select a partner, infer personality or decide what every Chinese household expects. A cultural label does not validate those decisions.
Find your path
Use the product map and its explained reading routes, rather than treating the milestone as a required reading list.

Checks with a defined scope.

Verification is repeatable work, not a hand-written row of green badges. Each local command records its outcome; browser evidence must cover the actual built routes and match the artifact fingerprint. Missing, failed or stale evidence cannot satisfy the final technical gate.

Tests and static generation

Run the complete web suite, the backend suite and a fresh static build. A successful build alone is not source approval or evidence that an optional remote service is available.

Discovery and metadata

Check the actual sitemap against independent routes, canonical URLs, robots rules, page titles, social assets, structured data and internal links. Search remains noindex.

Accessibility and responsive behavior

Scan the built sitemap routes and search at 320px and 1440px, record axe violations and incomplete checks, and inspect the new pages at 390px as well. Automated results do not establish WCAG conformance or a human screen-reader pass.

Performance and visual continuity

Keep the existing route-level resource budgets and inspect the established desktop/mobile screenshot cases. Lab budgets are not claims about field performance on every phone or network.

Sources and editorial state

Validate reference metadata, cultural-language checks and the bounded 20-page review ledger. Bind command and browser evidence to the actual generated artifacts; keep absent human approvals visibly pending.

The dated verification record lists actual command results and coverage. This page explains the rules; it does not turn an old lab result into a live performance or availability monitor.

Sources checked is not the same as human-approved.

The core review dated 2026-10-22 covers 20 selected pages and 35 active sources. It records 0 pages needing source-claim follow-up and 20 needing a human content sign-off. Selection followed user journeys and cultural-risk questions, not a claimed traffic ranking.

The source pass was AI-assisted and claim-scoped. Actual source text was distinguished from challenge pages and empty responses. Two unread original references were replaced with inspected institutional evidence; their failed retrieval history remains in the review ledger. The corrections include separating named race-story variants and two different Dunhuang catalogue references.

Approval of an implementation plan is not approval of its finished language. A human approval needs an identified reviewer, date, explicit content-sign-off evidence and the content fingerprint it covered. Changed content needs another check; neither build scripts nor this page grant approval.

Pages awaiting recorded human content review (20)

For an original reference rather than our summary, use the source directory. For what changed and why, read the corrections and review log.

What this milestone does not establish.

  • The calendar tools implement their documented conventions. Historical jurisdictions, a different year boundary or legal-age requirements may need another source. Read the methods and ranges.
  • Festival tables describe traditional observance dates, not official leave or local opening hours. A rolling window is not automatically a new independent verification.
  • Automated accessibility and lab budgets are bounded observations. They do not replace assistive-technology listening, physical-device checks, field data or every browser/network condition.
  • Optional almanac prose may be unavailable. Its fallback and cultural labels are not personalized forecasts; a model provider has not approved the library.
  • Image rights are item-specific. A museum credit is not a blanket license for the site. Check citation, image and reuse guidance.
  • Human review is unfinished. This remains a candidate rather than a claim that all 100 completion standards have been satisfied.

Read the dated record. Keep the caveats.

The machine-readable record retains actual test, build, metadata, accessibility and source-gate outcomes for a local snapshot. It identifies whether technical checks passed and how many human reviews remain; it does not assert a production release or official certification.

Open the dated verification record (JSON)

If the record says “not yet verified”, or does not match the artifacts being assessed, it cannot be used as a pass. Future builds need fresh command and browser evidence, not a copied success label.

Return to the useful routes and transparent progress report, or send a precise correction or review inquiry.