Selected records, deterministic tracing, human review
Financial discovery often fails before anyone reaches a legal conclusion. PDFs, spreadsheets, and narrative summaries drift apart. A balance on one tab does not match the statement page. A deposit total in a memo does not reconcile to the ledger. By the time counsel spots the gap, review hours are already spent and the exception arrives late.
Exit Protocol separates jobs that are often collapsed in manual review: record preparation, reviewer decisions, deterministic analysis, and professional interpretation. The broader platform includes production and source provenance, account coverage, ReviewedTransfers, Money Flow, deterministic LIBR, and conserved Value Lineage. This paper focuses only on the current single-claim LIBR workpaper contract. The export is attorney-reviewable. It is not legal advice and not an expert report.
Each workpaper preserves visible strategy, a material transaction ledger, source provenance for supporting rows, and SHA-256 snapshot integrity so a reviewer can inspect what was calculated, what was assumed, and what still requires professional judgment.
What this paper is, and what it is not
This paper is
- Technical product architecture for V1 LIBR workpapers
- A methodology explanation for deterministic tracing
- A description of the attorney and expert review workflow
- A V1 scope statement for single-claim exports
This paper is not
- Legal advice or a substitute for counsel
- An expert opinion or forensic certification
- A court filing or filing-ready work product
- A promise that output will be accepted in any proceeding
- A Clio approval, partnership, or certification claim
What a V1 LIBR workpaper exposes
The synthetic demo workpaper shows the review surfaces counsel and forensic accountants should expect before relying on any export. No screenshot is required to understand the structure: the packet is designed to make assumptions and evidence visible on the face of the document.
Visible workpaper components
Recorded-order, deposits-first, or withdrawals-first same-day mode disclosed on the workpaper face as a calculation setting, not a legal posture.
Material transaction rows tied to the replayed separate-property balance.
Supporting rows linked back to the selected statement or export used in review.
Cryptographic fingerprint of the calculation snapshot at export time.
Post-export file hash for integrity checks against later alteration.
Structured output for counsel and retained experts — not an automated legal conclusion.
One separate-property claim per account export in the current release.
Why reviewable financial intelligence matters
High-conflict asset disputes generate records faster than most teams can organize them. Statements arrive as PDFs. Adjustments live in spreadsheets. Explanations appear in email threads. Each format can be internally consistent and still impossible to audit as a single financial story.
Serious review depends on traceability. A reviewer must see which transactions moved a balance, which assumptions were applied, which sources supported each row, and where the record remains incomplete. The current LIBR workpaper makes that calculation chain visible rather than burying it inside manual tabs or generated prose.
Where financial discovery breaks down
Manual discovery workflows often fail for structural reasons, not because reviewers lack skill. The same failure modes appear across family-law, forensic accounting, and litigation-support engagements:
- Scattered PDFs: statements, tax forms, and payment-app exports sit in folders without a shared ledger spine.
- Spreadsheet opacity: hidden rows, pasted values, and one-off formulas make later audit costly.
- Provenance drift: conclusions detach from the exact page, account, date, or transaction that supports them.
- Late exception discovery: missing periods, duplicate entries, and unreconciled balances surface after narrative drafts are already circulating.
- Narrative without traceability: persuasive writing can outrun the underlying calculation path.
In adversarial review, credibility follows the chain of reference. The question is whether a reviewer can walk from a claimed remainder back to records, assumptions, and unresolved gaps. When that chain is weak, even accurate facts become expensive to defend.
Operating rules for credible forensic software
Exit Protocol is built for attorneys, forensic accountants, integration reviewers, and validation partners who need conservative tooling. The product rules below are intentional boundaries, not marketing language.
Deterministic math before narrative
Calculations replay from stated inputs and assumptions. Generated text cannot silently change balances or trace results.
Source-linked analysis
Material ledger rows retain a path to the selected statement, export, or supporting document used in review.
Visible assumptions
LIBR strategy, claim inputs, and methodology appear on the workpaper face so reviewers know what was applied.
Human review as product boundary
The export is attorney-reviewable and expert-reviewable. The platform does not replace professional judgment.
Uncertainty is preserved
Missing data, OCR ambiguity, unmatched transfers, and unresolved entity questions remain visible instead of being smoothed away.
Auditability and integrity
Access boundaries, audit logs, snapshot hashes, and file verification support disciplined record handling.
From selected records to a reviewable LIBR workpaper
V1 processing is deliberately staged. Record preparation, deterministic tracing, and narrative review stay separated so each step can be inspected independently.
1. Ingest
Accept selected bank statements, native exports, and supporting financial records. Preserve originals and create source references.
2. Normalize
Structure dates, descriptions, amounts, account identifiers, and running balances into consistent tables.
3. Reconcile
Test opening balance, deposits, withdrawals, and closing balance before professional reliance. Disclose unresolved controls: reconciliation is not a universal gate that prevents a tracing calculation.
4. Trace
Replay transaction history and apply selected LIBR assumptions across deposits, withdrawals, and balance dips.
5. Flag exceptions
Surface ambiguous counterparties, unmatched transfers, reviewer-expected periods not represented in the current record, and other review-required states.
6. Assemble workpaper
Export an attorney-reviewable LIBR workpaper with strategy, material ledger rows, provenance, and integrity hashes.
LIBR tracing is chronological state, not summary logic. The system replays transaction history to apply selected LIBR assumptions. Later deposits do not automatically restore prior separate-property trace value unless applicable facts and legal theory support that treatment. That constraint is why deterministic replay matters and why V1 keeps the calculation path visible on the workpaper.
LIBR workpaper layers and review boundaries
The broader platform also includes Document Production, account coverage, ReviewedTransfers, Money Flow, and Value Lineage under separate current contracts. The table below describes only the specific current LIBR workpaper path; each layer has a defined responsibility and an inspection point for counsel, forensic accountants, or technical reviewers.
Swipe to view all columns on tablet and phone widths.
| Layer | Function | Review Value |
|---|---|---|
| Evidence Intake | Accepts selected records, preserves originals, and creates source references. | Maintains chain-of-reference between output and source material. |
| Normalization | Structures financial data into consistent transaction and account tables. | Reduces manual re-entry and enables reconciliation. |
| Reconciliation | Tests balances, duplicates, missing periods, and file inconsistencies. | Surfaces weaknesses before tracing outputs are drafted. |
| LIBR Tracing Engine | Replays account history and applies selected LIBR assumptions deterministically. | Produces reproducible trace results that can be re-run and challenged. |
| Report / Workpaper Assembly | Builds the attorney-reviewable export with strategy, ledger, provenance, and methodology. | Turns calculation output into inspectable review material. |
| Verification / Integrity Layer | Captures snapshot SHA-256 values and final file hashes for post-export checks. | Supports integrity verification without claiming absolute certainty. |
Assistance without silent authority
Generative models can help with drafting, classification, summarization, and review prompts. They are not reliable authorities for balances, ownership, credibility, courtroom acceptance, or legal entitlement.
Exit Protocol keeps assistive AI separate from authoritative financial state. Extraction or AI tools may prepare editable proposals where explicitly configured, but they do not create reviewer decisions, calculate LIBR, finalize Value Lineage, or decide ownership, credibility, admissibility, or legal entitlement. Deterministic calculations and reviewer-controlled records remain authoritative.
Disciplined handling for sensitive financial records
Financial-dispute records may include account numbers, tax data, business ownership details, and litigation strategy. Security posture is part of product credibility, not a separate IT brochure.
- Access boundaries: separate public demo surfaces from authenticated case workflows.
- Audit logs: record important actions so file handling can be reviewed after the fact.
- SHA-256 snapshot integrity: fingerprint calculation state at export time.
- Final file hash verification: compare exported bytes against later copies.
- Data minimization: process selected records required for the review task.
- Review trail: preserve enough context for professional inspection without overstating guarantees.
These controls reduce operational risk. They do not eliminate legal risk, human error, or the need for independent professional review.
Realistic adoption paths
Teams evaluate Exit Protocol at different speeds depending on matter sensitivity, technical review requirements, and integration needs. The common path starts with bounded proof, then moves to professional review on live records.
Swipe to view all columns on tablet and phone widths.
| Mode | Best Fit | Purpose |
|---|---|---|
| Sandbox Evaluation | Attorneys, forensic accountants, and legaltech reviewers. | Inspect workflow using synthetic records without private client data. |
| Professional Review | Law firms and forensic professionals on active matters. | Organize selected records and produce attorney-reviewable LIBR workpapers. |
| Clio Selected-Document Workflow | Firms already operating inside Clio matter management. | Authorize access, choose a matter and source document, and stage extraction candidates for explicit local review. No automatic workpaper return is implemented. |
| Partner Integration | Legaltech, accounting, and financial-data platforms. | Connect structured transaction or document data to tracing and workpaper outputs. |
| Controlled Deployment Environments | Organizations with heightened data-boundary requirements. | Support firm-specific infrastructure, security review, and deployment constraints. |
Explicit product boundaries
A credible forensic platform states what it does not do. Exit Protocol prepares structured review material. It does not determine legal rights, replace counsel, replace retained experts, or convert incomplete records into certainty.
- V1 scope: workpaper exports are scoped to single-claim accounts in the current release.
- Multi-claim matters: The current workpaper does not provide multi-claim allocation; such matters require a separately reviewed methodology and workflow.
- Value Lineage boundary: Current modeled source value remains inside one account and may reach a reviewed outgoing transfer boundary. It does not create destination TraceLots or cross-account source-value propagation.
- Parsing risk: OCR and parsing can introduce errors that must be reconciled against source documents.
- Jurisdiction dependence: LIBR is jurisdiction-specific and fact-dependent.
- Entity resolution: suggested relationships may require independent confirmation.
- AI drafts: AI summaries are drafts, not findings.
- Residual risk: Security controls reduce risk but do not eliminate legal, operational, or human risk.
- Professional review: Counsel and retained experts must validate assumptions, jurisdiction, and use before any filing or testimony.
The output is attorney-reviewable. It is not legal advice and not an expert report. It is not a court filing.
Better review material, not automated legal conclusions
Exit Protocol is designed to make financial review more organized, reproducible, and challengeable. This paper's single-claim LIBR workpaper gives attorneys, forensic accountants, and auditors a visible calculation path, a material ledger, source provenance, and integrity hashes they can inspect before relying on that export. Reviewed transfers, Money Flow, and Value Lineage are separate current platform contracts.
The goal is not automated legal conclusions. The goal is better review material: selected records converted into structured workpapers that respect professional boundaries, preserve uncertainty, and stand up to adversarial scrutiny. Evaluate the synthetic demo, review the methodology, and decide whether the workflow fits your matter under independent professional judgment.