Limits and Verification

Plurilore narrows an AI answer to retrieved source material and exposes that material for inspection. Those controls make verification possible; they do not make generated prose infallible. The model can miss context, combine distinct claims, misread a translation, or sound more certain than the evidence allows.

Verify the source chain

1. Open the Sources consulted panel.
2. Confirm that the cited work and passage match the question.
3. Read before and after the highlighted passage in the Library.
4. Separate direct statements from the model's inference.
5. Check whether an unrepresented source, school, translation, or historical context could change the conclusion.

For comparisons, verify at least one relevant passage from every selected tradition and keep differences visible. Similar vocabulary does not establish equivalent doctrine or practice.

Know the corpus boundary

Each agent searches only the works listed on its Wiki page. A corpus may contain one historical translation, selected scriptures, or a limited group of ethnographic texts. Absence from an answer can mean retrieval missed a passage or the necessary source is not indexed; it does not prove that a tradition has no teaching on the subject.

Do not use Chat as an authority

Plurilore is not clergy, a religious court, a therapist, a doctor, or a legal adviser. Do not rely on it for religious rulings, crisis support, diagnosis, treatment, legal decisions, or individualized pastoral direction. Those questions require qualified people and appropriate primary sources.

Protect sensitive material

Chat is ephemeral rather than a saved server-side history, but live requests still require processing. Do not submit secrets or highly sensitive personal information. Notes and private Practice reflections are local to the current device; exported files inherit the sensitivity of their contents. See the privacy policy for the full data-handling terms.