CREB™ Summary Report - Court Ready Evidence Bundle - Executive Summary.

Court Ready Evidence Bundle

Evidentiary Summary Report

Bundle ID: CREB™-PS-2026-04-19-A8F3C2D1 · Issued: 2026-04-19 13:47:22.108 UTC

1Summary Overview

This Court Ready Evidence Bundle (CREB™) provides a structured, verifiable, and traceable presentation of findings derived from authorized data sources for the Pre-Settlement Attestation pillar.

The CREB™ framework is designed to:

• Identify and document material anomalies, inconsistencies, or risk indicators
• Preserve evidence in a court-ready, audit-ready format
• Enable independent validation, reproducibility, and review
• Support compliance, investigative, regulatory, or legal workflows

All findings are supported by source-attributed data, structured analytical methodologies, and cryptographic validation with chain-of-custody controls. This report presents evidence-backed observations, not final legal conclusions.

2Purpose

The purpose of this CREB™ is to:

• Provide a standardized evidentiary representation of identified patterns or irregularities in the subject pre-settlement transaction
• Enable authorized stakeholders (the requesting custodian, counsel, regulators on subpoena) to assess potential exposure, risk, or non-compliance
• Support further review, audit, investigation, or escalation, where appropriate
• Ensure that all findings are preserved in a manner suitable for formal proceedings, if required (FRE 902(14) self-authentication framework, asserted under JIL patent claim 53)

3Scope of Analysis

Subject

Pre-Settlement Attestation request from Custodian-X (VND-CX-0047) for outbound transfer to Recipient-Y (LEI 549300XXXXXXXXXXXXXX)

Time Range

2026-04-19 13:47:21.880 UTC (request received) → 13:47:22.108 UTC (verdict issued); analysis window 228 ms

Data Sources

OFAC SDN + Consolidated · OFSI · UN · EU · OIG LEIE · NPPES · POS file · CMS Inpatient DRG · OpenSanctions / PEP · GLEIF LEI · IRS NFTL · CourtListener · CMS Owners (PECOS) · OpenCorporates · ATTOM Property · CERT FY2024 detector rules

Customer Profile

profile_id prf_a3f9b1c84e6d50d2 (LOB: Custodian, Tier 1 + Tier 2 enabled, 87 of 148 catalog checks active)

Authorized Use

VND-CX-0047 onboarding agreement § 4.2 + per-attestation legal basis recorded in attestation envelope

Review limited to approved datasets and the defined time range. No unauthorized data access or external disclosure has occurred. Customer payment data never left the customer's perimeter; only metadata required to evaluate the 87 enabled checks was processed.

4Summary of Findings

The 87-check pre-settlement analysis identified the following categories of observations:

• Pattern-based anomalies: 2 findings - counterparty velocity anomaly (z = 3.4) and routing-account-change detection (modified within last 7 days)
• Entity or identity correlations: 1 finding - counterparty LEI links to a UBO 2 hops from a 2024 OFSI consolidated entity (graph distance 2)
• Behavioral or transactional irregularities: 1 finding - request volume spike (4.2× sender's 30-day median) outside operator hours (UTC 22:47 local)
• Statistical or operational outliers: 0 findings on monetary amount; 0 on geographic corridor; 0 on rail-specific (SWIFT gpi chain integrity, IBAN MOD-97)

These findings may indicate:

• Potential inconsistencies with the requesting custodian's standard outbound profile
• Elevated risk requiring additional validation before settlement
• Conditions warranting review by Custodian-X's compliance officer prior to release

Final verdict: REVIEW (confidence MEDIUM, z-score combined 3.4 across 4 corroborating signals). The custodian's policy decides whether to release the transfer; JIL attests to the findings, the custodian decides on the action.

5Evidence Characteristics

All findings supporting this CREB™ meet the following criteria:

• Traceability: each data point linked to a verifiable source with an attestation certificate ID, last-refresh timestamp, and update cadence (see Detailed Verdict Record § 3 for the source manifest)
• Reproducibility: analytical results can be independently replayed bit-for-bit from the reproducibility manifest in § 6 below
• Integrity: evidence protected against alteration via SHA-3-512 hashing at acquisition + Merkle-rolled into the bundle root
• Attribution: source origin preserved and documented for every check (vendor-approved data sources, JIL-internal data sources, customer-provided context)
• Consistency: findings validated across multiple data points and datasets - single-source flags are escalated to MEDIUM confidence; multi-source corroboration drives HIGH confidence

6Chain of Custody & Validation

This CREB™ carries a tamper-evident chain of custody, including:

• Cryptographic hashing of every check input + output at time of acquisition (SHA-3-512)
• Timestamped ingestion + processing records (microsecond-precision event log; see Detailed Verdict Record § 7)
• Immutable logging of every transformation and access event in the JIL audit trail (postgres trust.audit_log, Merkle-chained on insert)
• Preservation of original source references (every data fetch's request URL + response hash + auth context recorded)
• Independent signing-node signature (Ed25519 + Dilithium-III hybrid) over the bundle root
• JIL evidence-ledger court-anchor transaction (block 9_837_241, tx 0xa4f9...c8e3)

This structure supports evidentiary reliability and is aligned with recognized standards for authentication, including FRE 902(14) self-authentication readiness per JIL patent claim 53.

Reproducibility Manifest

A third party with the manifest below + the same source data can replay this analysis and produce a bit-identical finding set:

profile_id            : prf_a3f9b1c84e6d50d2
profile_check_set_hash: sha256:7e2a91f0d83b6c5e4a17b9c3d8f0e1b2a4c6d7e8f0a1b2c3d4e5f6789abcdef0
code_version          : jil-attestation@2026.04.19-948 (sha 7d4e9c2)
input_hash            : sha3-512:b5f4a8e7d6c5b4a3...
query_plan_hash       : sha256:9f3a7e2d8c5b4a91...
source_data_version   : ofac-2026-04-19, oig-leie-2026-04-15, openSanctions-2026-04-19,
                          cms-inpatient-2022-23-fy24, cert-fy2024-rules-v1.4
signal_thresholds     : z_floor=3.0, multi_signal_min=2, high_band_z=4.0+3sigs
verdict_engine_seed   : deterministic (no stochastic LLM in verdict path)
ava_runtime           : in-house (no external LLM API call; agentic layer rode on top of finding set)
ava_llm_enabled       : false
ava_llm_model         : n/a (template-only)
ava_llm_calls         : 0 of 0 narratives refined; 0 of 0 remediation plans refined
ava_llm_gate_state    : LLM_ENDPOINT=set, LLM_ENABLED(env)=false,
                          engagement_parameters.llm_enabled=false, customer_profile.plan=poc
  

LLM gate (this CREB™): the in-house LLM was disabled for this engagement. All findings, narratives, and remediation plans were rendered from versioned reference templates only. When the gate is open (paid customer who has explicitly opted in), narrative + remediation copy passes through a refinement step on a JIL-controlled LLM (Llama 3.3 70B AWQ), with a numeric-drift guard that rejects any rewrite that mutates a $ amount, %, count, or date. All prompts and responses stay inside JIL infrastructure.

Replay command (for the customer's compliance officer or counsel):

jil-attest replay --bundle CREB™-PS-2026-04-19-A8F3C2D1 --manifest ... above ...

7Evidence Organization

All supporting materials in this CREB™ are organized within a structured evidence index. Each item has:

• Unique Evidence Identifier (EVD-PS-A8F3C2D1-001 through -087, one per executed check)
• Source reference (attestation certificate ID + dataset + record key)
• Acquisition timestamp (microsecond-precision)
• Integrity verification hash (SHA-3-512)
• Relevance classification (PRIMARY / CORROBORATING / OUT-OF-SCOPE)

This enables efficient review, validation, and downstream use by authorized stakeholders. The full evidence index lives in the companion Detailed Verdict Record.

8Risk & Significance Indicators

Based on the observed findings, the following considerations apply:

• Presence of multi-signal corroboration (4 signals fired across pattern, identity, behavioral categories)
• Indicators of moderate impact on this single transaction; counterparty has no prior pattern in the requester's 90-day history
• Potential misalignment with the requesting custodian's standard out-of-hours velocity profile and the documented routing-stability expectation

Risk classification: MEDIUM

(LOW = single-signal outlier · MEDIUM = z ≥ 3 with ≥ 2 corroborating signals · HIGH = z ≥ 4 with 3+ signals · CRITICAL = z ≥ 5 + sanction proximity ≤ 1 hop)

9Intended Use

This CREB™ is intended to support:

• Internal compliance and audit processes at Custodian-X
• Investigative review and validation by Custodian-X's compliance officer
• Regulatory or program integrity workflows where Custodian-X elects to share the bundle (OCC examination, NYDFS Part 500 audit, FinCEN inquiry, etc.)
• Legal or evidentiary preparation if Custodian-X faces challenge to the disposition of this transfer

Use of this report should be limited to authorized personnel and approved purposes. JIL Sovereign retains no copy of the underlying transaction data; only this CREB™ and its embedded hashes.

10Limitations

• This report does not constitute a legal determination or final conclusion; the action (release / hold / decline) is the custodian's decision
• Findings are based on the data sources and the 87 checks in Custodian-X's authorized profile at the time of analysis
• Additional data (a customer's internal KYC file, ongoing investigation context, recent SAR filings) may refine or alter interpretations
• Independent review is recommended prior to any formal action (regulatory filing, civil litigation, contractual escalation)
• Out-of-scope checks (61 of the 148 catalog) were not run because Custodian-X's profile excludes them; their absence does not imply a finding either way

11Next Steps (Optional)

Authorized stakeholders may consider:

• Targeted review of the 4 individual findings (drill-down available in the Detailed Verdict Record § 4)
• Tier 2 substantiation if Custodian-X wants UBO depth-3 graph traversal, premise verification, or commercial-data enrichment on the counterparty (escalation cost: per-case under the customer's Tier 2 bundle)
• Internal escalation within Custodian-X's compliance organization (e.g., to the SAR review queue)
• Engagement of compliance, investigative, or legal resources if the elevated risk and counterparty proximity-to-sanctions trigger the custodian's policy threshold

END OF SUMMARY. This document is the executive summary of CREB™-PS-2026-04-19-A8F3C2D1. It is paired with the Detailed Verdict Record (9-section operational data: identifiers, subject, test profile, approved-checks matrix, out-of-scope checks, final verdict, cryptographic seal, chain of custody timeline, authorized evidentiary uses). Both ship together as one CREB™. Subject data in this sample is fictitious. Verdict-engine architecture: in-house Agentic AI + Data Mining; no external LLM API in the inference path. Patent-pending architecture under JIL Sovereign Holdings, LLC (Wyoming) provisional claim filings 49-53.