What works. What remains open.

1. Contestation: email intake and human review

The challenge form creates a reviewable local dossier and opens a message to dataroom@csrholding.eu. Attach the downloaded dossier and send it yourself. Receipt requires a reply from the human reviewer; this site has no ticket database or delivery tracking. Reconsideration uses the same reference. An independent appeal body and response times are not yet established.

2. v0.14 theory: coherence, reality and pattern challenge

The public foundation now distinguishes internal coherence, inferential validity, true premises and correspondence with reality. CE-15 through CE-18 remain candidates. The six-step pattern challenge accepts bounded corroboration, rejection or insufficient evidence; failed controls cannot be hidden by an average score. “AI pareidolia” remains an explicit metaphor, not a diagnosis.

3. Sources: explicit coverage, unfinished validation

The source register separates read references, supported claims, scope limits and specialist validation. All 197 atlas records have a review state. Source coverage is partial; no specialist approval is invented.

4. Languages: translated atlases and clearer scope

The English atlases retain editorial translations of all 182 epistemic entries, 15 cosmogonic clusters and 12 patterns, preserving the Italian v0.10 atlas baseline. The theory pages now reflect v0.14. Web essays remain editorial readings, not certified line-by-line translations of the full documents.

5. Downloads: complete v0.14 reading formats

Both Italian originals are available as DOCX and as PDF, EPUB, HTML and Markdown exports, with a checksum manifest. The v0.10 documents remain in the historical archive and the older English v0.4 PDF remains clearly labelled.

6. Application: a working case worksheet

The guided case lab separates evidence, inference, uncertainty, alternatives, values, affected people and the risks of action and inaction. It flags missing fields and exports a dossier. It does not consult an AI, evaluate truth, authorize action, certify independence or pass the nine constitutional Gates. The validated Oracle remains a research objective.

Next evidence needed

• Named specialist and relevant community review of atlas claims, translations and the v0.14 candidate protocol.
• A documented record of human intake, reasoned responses, conflict management and actual reconsideration.
• Evaluation of pattern challenges and the case procedure on documented cases, including adverse outcomes, before claiming Oracle effectiveness.