Definitions for both macro sets in OuiDire.app, plus the core auditability and workflow terms around them.
Second-hand claim without primary support in the cited segment.
A claim attributed to someone else (“X said…”) without direct observation or primary documentation in the cited segment. Treat as a provenance flag: who said it, when, and with what evidentiary basis?
Warping patient speech (meaning/tone/implication).
The record claims to reflect the patient’s words or intent but shifts meaning, tone, or implication via selective paraphrase, reframing, or insinuation. Key signal: divergence between attributed speech and asserted proposition.
Weak cues → strong claims, without intermediate evidence.
Invented events or unjustified inferences presented as factual. Includes “stretching” from weak cues to strong conclusions (risk, intent, diagnosis) without intermediate evidence.
Complex life reduced to a durable institutional storyline.
A life history is invented, or condensed into a durable storyline that becomes institutional “truth,” often by omitting counterevidence, flattening nuance, or reinterpreting past events to fit a diagnostic arc.
Past labels become present evidence through repetition.
Reusing past labels or narratives as current evidence, without re-verification. Creates narrative lock-in: earlier claims gain authority through repetition rather than fresh corroboration.
Material contradictions (not minor details).
Incompatibilities within a document or across the set (dates, sequence, observations, risk claims). Not minor typos: contradictions that change meaning, interpretation, or credibility.
What’s missing changes the meaning.
Missing context or counterevidence a reasonable reader would expect, where absence materially shifts interpretation (procedural facts, alternative explanations, prior corrections, exculpatory elements).
The claim escalates without proportional evidence.
When wording, qualifiers, or details escalate in intensity across the text (or across documents) without proportional new evidence: tone hardens, scope expands, or certainty increases. This is a **mechanism macro** (not a Source tag). Typical cues include connectors like “therefore,” “clearly,” “without a doubt,” or generalizations that outrun the cited segment.
Narrative repetition: a false story hardens into “fact.”
When a story/claim (often false or never revalidated) propagates across documents (“as previously noted…”), hardens, and ends up treated as a fact without revalidation. Unlike amplification, canonisation focuses on **propagation** and **stabilization** of the narrative (prior-record references), not only tone.
Objective checks exist, but are set aside.
Available objective testing, measurement, or verification is declined while a strong interpretive claim is maintained.
Competing explanations are not meaningfully compared.
Plausible alternative explanations are available but are not meaningfully examined before a preferred explanation is retained.
A single cause closes the reading too early.
A single explanatory frame is imposed too early, preventing competing causes from being seriously considered.
Treatment effects are minimized or sidelined.
Medication effects or treatment-induced distress are acknowledged only in part and are not fully integrated into the interpretation.
Risk is overstated even though uncertainty is explicit.
Serious restrictive responses are justified despite explicit uncertainty about actual danger or risk.
Trauma is acknowledged, then pushed to the background.
Trauma is recognized but deferred or subordinated in favor of a dominant diagnostic label.
Ambiguity is turned into leverage.
Uncertainty, vague concern, or indeterminate risk is used to support pressure or threat-based leverage.
The surface is managed while the underlying cause stays untouched.
Immediate symptoms or behaviors are managed while underlying causes are acknowledged yet left unaddressed.
The precise passage a reader can review and verify.
Precise passage from the source text used to support a suggestion, macro, or claim. It is what the reader reviews to verify whether the result is actually supported by the file.
The stable link between a result and its source.
Stable reference that ties a result back to its original source: document, page, and text region when available. It should remain valid in the interface, in exports, and after reformatting.
Not strong enough to support the conclusion.
Excerpt that is too short or too weak to support a useful conclusion. For example, a few isolated words may not be enough to support a macro or a claim.
Expand the excerpt without losing the source link.
When an excerpt is too thin, OuiDire can expand it to the full sentence or a nearby bounded window to make it more understandable. The source reference is still preserved so traceability remains intact.
Fast lane on the current macro set.
OuiDire's fastest AI pass. Call Gemini runs Gemini on the current macro set to surface more suggestions card by card. It is the fast lane when you want to unlock proposals without changing your reading logic.
A deeper GPT second pass with exportable findings.
Slower, deeper GPT analysis organized around 30 micro-tags derived from the macro-set logic. It does not replace that logic: it pushes it to a more granular level once the main annotations are already well advanced, then produces detailed exportable findings.
Deep lane across both macro sets.
Wider AI pass that runs Gemini across both macro sets, then gathers the results in Gemini Digest card by card. Use it when you want to compare two readings or surface many more suggestions.
Gemini results, exports, and direct return to the card.
Results surface for Gemini passes. It gathers suggestions, gives access to exports, and lets the user jump straight back to the current card for review.
See what is happening (OCR/AI logs + storage/balance).
A technical section inside OuiDire that makes the workflow visible (OCR logs, AI call logs, storage/balance status): progress, steps, errors, and current state. Goal: reduce opacity when an action takes time and make troubleshooting easier.
From card level to the wider analytical terrain.
Zoomed-out surface that moves from the card level to the wider terrain: human signal, dominant macro pattern, and strongest intersections. Map-3 becomes useful after several annotations, once document-level patterns start to emerge.
From source pages to usable cards.
OuiDire turns PDFs and document photos into cards through an OCR pipeline designed around pages, structure, and traceability. Depending on the ingestion path, the system can rely on Google OCR or Azure OCR. The goal is not raw text alone, but extraction that remains usable for cards, review, and source citations.
Fast engine for Gemini suggestions.
Preferred Gemini model for fast suggestions in Call Gemini and Double Run. It is chosen to keep a good speed/cost balance on card-by-card analysis.
GPT engine for the deeper critical pass.
OpenAI model used for Deep Critical Read. This layer remains tied to the macro-set logic, but pushes it further through 30 micro-tags derived from that reading grid. It is meant for heavier, slower, more detailed, exportable analysis.
Review before prod, commit by commit.
Per-commit preview URLs used to review changes before production. Enables fast iteration and safe review of content and UX updates.