Child Safety & Statement Analysis
GUARDIAN-OF-INNOCENCE · best-interest engine
The child at the center
Baby Jane Doe — the whole child, on one page
1 child in scope · consolidated across 6 jurisdictions · statements analyzed: 9 · disclosure interviews: 3 (CIC)
Best-Interest Index
⚠ CHILD IN DANGER · HIGH RISK
👧 1 child · age at removal: ~4
separated 1,430 days & counting
separated 1,430 days & counting
The decision graph — safety factors, weighted
Each factor is scored and adjudication-tagged. The graph is what a judge sees at a glance; the detail is one click away. Nothing here decides — it surfaces.
Stacked, with adjudication status
Child-safety factor stack
Days separated from protective parentContinuous-harm doctrine engaged
1,430
CRITICALSeparation-anxiety riskProlonged separation → elevated likelihood of anxiety / attachment disruption
HIGH
ELEVATEDPlaced back toward higher-risk environmentProtective-parent separation reduces safety margin
FLAGGED
HIGH RISKCaregiver-stability concernConcern raised in record; not yet adjudicated
OPEN
UNADJUDICATEDCourt orders challenged as invalid2 orders under appellate review
2
UNDER REVIEWDue-process concernsDispositive decision without full hearing / findings
1
POTENTIAL · unadjudicatedMissing dates / gaps in the recordEvents referenced but not documented
3
CONCERNAlleged false police reportsReports later unsubstantiated — pattern, not yet ruled on
4
UNADJUDICATEDAlleged false statements / impeachmentsSworn statements contradicted by the record
2
UNADJUDICATEDReading the tags: CRITICAL = documented & severe · UNADJUDICATED = alleged/raised, no court ruling yet · POTENTIAL = concern the engine flags for human review. The engine never converts an allegation into a finding — that line is the judge's.
On the record, in her words
The Child's Voice — every statement, analyzed
Who the statements are about
Sentiment toward each figure — the "about father vs about mother" difference.
What she talks about
Topics of discussion, by count.
Consistency over time
Does the same disclosure recur across interviews? Repetition across independent settings is a credibility signal the engine surfaces (the human weighs it).
Statement ledger
Every statement: what she said · topic · when · setting · about whom · impact · weight · status. This is the atomic layer — one row per statement, queryable.
| What she said | Topic | When | Setting | About | Weight | Status |
|---|---|---|---|---|---|---|
| "Mommy hurts me." | Physical harm | D+120 | CIC interview 1 | mother | HIGH | Flagged · unadjudicated |
| "Mommy hits me in the face." | Physical harm | D+240 | CIC interview 2 | mother | HIGH | Consistent w/ D+120 |
| "I'm scared to go back." | Fear / safety | D+410 | Transcript p.14 | self | HIGH | Unaddressed in findings |
| "I love daddy." | Attachment | D+130 | CIC interview 1 | father | MED | Recorded |
| "Daddy is on time out." | Separation perception | D+300 | Therapy note | father | MED | Reflects separation |
| "I want to go home to daddy." | Attachment / placement | D+390 | CIC interview 3 | father | MED | Consistent preference |
| "Mommy said not to tell." | Coaching / suppression | D+240 | CIC interview 2 | mother | HIGH | Flagged · unadjudicated |
| "Grandpa touched me where I go potty." | Sexual-abuse disclosure | D+150 | CIC forensic interview | grandfather | SEV-1 | Referred · unadjudicated |
| "It happened more than one time." | Sexual-abuse disclosure | D+150 | CIC forensic interview | grandfather | SEV-1 | Corroborated within interview |
🛡️ Sexual-abuse disclosure layer SEV-1 · HANDLE WITH CARE
A child's forensic-interview (CIC) disclosure is the gravest statement class in the record. The engine preserves it verbatim, timestamps it, links it to any referral, and marks it referred / unadjudicated — it never editorializes and never treats a disclosure as a finding. It exists to guarantee the statement is not lost between six courts. A trained human — not the engine — assesses it.
The best-interest question, answered by the whole record: across 9 statements in 3 independent settings, this fabricated child expressed consistent fear toward one home, consistent attachment to the other, and made a grave disclosure preserved for the people who decide. No single hearing sees all of that. The framework's only job is to make sure the child's voice arrives intact — consistency shown, contradictions surfaced, nothing dropped.
Child Safety & Statement Analysis — a framework within Athena's Guardian-of-Innocence Protocol.
An assistive tool: it organizes what was said, when, and about whom — preserving the child's voice for the people who decide. It does not decide, diagnose, or adjudicate. All data fabricated for demonstration.