Freedom of inquiry should be observable in practice.

For independent publishers and researchers

1. Publish original work and provide stable URLs, descriptive titles, dates, authorship and correction notes where feasible.
2. Link to primary evidence and meaningful criticism, including relevant perspectives with which you disagree.
3. Make reference lists, feeds and source collections accessible independently of a dominant search service.
4. Describe whether a cited item is a supported finding, a personal opinion, an unresolved question or an outlier hypothesis.
5. State permitted uses of your work; distinguish search indexing, retrieval access and model-training rights without promising that publication forces AI inclusion.
6. Invite reasoned rebuttals and corrections without conditioning access on ideological agreement.

For search engines and AI developers

1. Audit coverage at each stage: crawling, indexing, training, filtering, retrieval, generation and citation.
2. Evaluate performance across languages, regions, cultures, unusual source types and legitimate minority perspectives; publish scope and limits.
3. Differentiate source inclusion from evidence endorsement. Explain how disputed or weak claims are handled.
4. Give users routes to inspect original sources and select useful discovery modes, including broad-context and competing-arguments views where feasible.
5. Document material restrictions, source-removal categories, corrections, and proportionate review or appeal options.
6. When declining a request, explain the specific unsafe capability or legal boundary rather than attributing a belief to the questioner.

A reproducible six-step discoverability audit

1. Define the claim: identify the specific source, proposition, query, expected result and why it is relevant.
2. Check access: verify the URL, HTTP response, indexability, canonical URL, language, robots instructions and publication date where observable.
3. Repeat the test: record date, location or locale, system, mode, exact queries, neutral paraphrases and comparison queries.
4. Record what appears: separate missing links, poor ranking, inaccurate summaries, refusals and absent citations.
5. Test alternatives: compare other engines, direct URLs, links on independent sites, and plausible quality or policy explanations.
6. Seek correction: keep screenshots or observations, use any relevant provider tools or appeal route, and date subsequent results.

The viewpoint workbench helps visitors export a documented comparison without submitting their notes to this website. An individual negative result does not establish a coordinated suppression policy.

Suggested measures of a genuinely open discovery experience

For all measures, publish the sample selection, the systems and dates tested, limitations, and negative findings as well as positive findings. These are suggested research measures, not established legal obligations or validated universal scores.

A first 90-day implementation sequence for this website

1. Days 1–30: publish and link the expanded charter; verify accessible sources, record page and content ownership, and invite editorial feedback.
2. Days 31–60: build annotated cross-site research trails on multiple topics, include documented criticism, and test navigation with varied readers.
3. Days 61–90: pilot repeated searches on narrow, testable questions; publish discoverability observations, corrections, and any limitations.

The proposed schedule is a plan, not a claim that visitor studies, provider audits or partnerships have already occurred. The site remains an un-deployed 1.38.0 WIP source artifact.

Our non-negotiable distinction

Start at the full expanded charter, explore the independent source trails, or share your own research.