Noor described mood episodes, periods of surveillance certainty outside obvious mood change, and a contractor harassment campaign.
Open Noor Devlin dossierINTERNAL / LONGITUDINAL
GLASS CATALOGUE Longitudinal Review
A multi-year review of mood, psychosis, work, harassment, recovery, and copied institutional language in Noor Devlin’s record.
EVENT RECONSTRUCTION
Chronology by source.
Entry order is UTC. Reliability measures source integrity, not moral truth.
Noor’s strongest published findings predate the admission and remain reproducible.
Open Elias Quill dossierThree clinical summaries copy phrases from a contractor threat assessment rather than Noor’s own words.
Open Imani Vale dossierDiagnostic uncertainty remains. Mood, psychosis, functioning, and factual harassment require separate longitudinal lanes.
Open Dr. Laila Mensah dossierNoor identified two conclusions they now consider overextended and requested they remain visible as revisions.
Open Noor Devlin dossierSERVER-SIDE REDACTION
Additional event entries sealed
The current case is visible, but some source-specific entries require a higher compartment.
1 sealed record omitted from this response.
Open access terminalHUMAN STAKES
A case conclusion changes people, not only records.
A multi-year review of mood, psychosis, work, harassment, recovery, and copied institutional language in Noor Devlin’s record. The case changes people’s safety, credibility, relationships, work, care, and future choices; it is not only an evidence puzzle.
Dignity risk
Treating diagnosis or distress as proof of deception Treating access as proof of action
False-inference risk
Treating diagnosis or distress as proof of deception Treating access as proof of action Treating a confident institutional record as neutral fact Treating one mistaken claim as global unreliability Treating cooperation as loyalty or refusal as guilt
Immediate harm
That every accurate finding will be dismissed and every mistake archived forever
Long-term cost
Leaves handwritten source questions inside library books, then returns to remove them. Wants to finish a quiet book about municipal corruption that contains no secret programs and endangers nobody. Wants to teach analysts how to document uncertainty without using it to avoid moral responsibility. Leaves one evening each week unscheduled to protect a life not organized by crisis.
Decision threshold
Act only at the consequence threshold supported by corroborated evidence; uncertainty is not permission to abandon the person or convert suspicion into a verdict.
Humane player options
- Separate observation from interpretation before escalating.
- Preserve the person’s ability to contest and correct the record.
- Use the least harmful reversible action consistent with immediate safety.
- Document uncertainty, counterevidence, and institutional incentives.
Debrief unresolved human questions
- Who was harmed by the investigation process itself?
- What information, choice, work, housing, care, or relationship must be restored?
- Which conclusion should remain provisional?
- Who can contest or correct the record?
EVIDENCE LAB
Methods for testing this case without collapsing uncertainty.
Each lens separates observations, claims, source quality, alternative explanations, and consequence thresholds. The exercises are fictional, untimed, and do not collect answers.
PERSONA CROSS-EXAMINATION
The case tests several different persona mechanisms.
These links explain how role, access, loyalty, current state, evidence behavior, and relationship function change the player’s reading of this file.