Skip to content
Agent packs

Disputes · Agent pack

New pack

Card dispute adjudication

Cadre pack · Arbiter

Tracks the supplied timing rules and evidence tests before returning an internal disposition for accountable review.

Demonstrated on representative cases. Source coverage, thresholds, policy interpretation, escalation design, and performance require institution-specific validation.

Operating dossier

The job, permitted decisions, and owner

Mandate

Separates configured Reg E and Reg Z timing states from the underlying dispute merits, preserving both for accountable review.

Case arrives with
Dispute ID · Now ts
Allowed outcomes
Structured outcome for review
Where it fits
Fraud & Disputes
Decision logic
  1. 01Organizes a cardholder dispute using the supplied Reg E or Reg Z policy configuration without collapsing timing state into the underlying merits.
  2. 02Checks the configured clock before the merits and records any interim-credit action identified by that policy while the evidence review continues.
  3. 03On the merits it applies the evidence tests in the supplied policy: address and card-code results, delivery evidence, prior undisputed use, and matching device or network context inform the review without deciding it automatically.
  4. 04Records the interim action, timeline status, and merits status separately. A ruling or liability party appears only when the evidence supports one; the record itself moves no money.
Must not
Invent missing evidence or act outside the permitted outcomes.
Human owner
Conflicting evidence and exception handling
Evidence retained
Structured outcome, observed path, and the source evidence used.

Case and record

One case, from supplied evidence to recorded decision

Supplied case

Representative case dataSynthetic or simulated data
Dispute ID
D1
Now ts
2026-06-24
View the case data

Case key clear-unauthorized-chargeback

{
  "disputeId": "D1",
  "nowTs": "2026-06-24"
}

Evaluation and limits

Evaluation coverage and stopping points

Published evaluation map

Expected decisions, boundaries, and costly failure modes.

These cases show what the pack is asked to decide, when it should stop, and which plausible errors the evaluation is meant to expose. They are not a reliability score or independent validation.

Expected and borderline behavior

Routine judgments, close calls, and named stopping points.

  • Reg E card-not-present fraud, AVS+CVV fail, no merchant evidence

    chargeback, cardholder wins.

    Case key

    clear-unauthorized-chargeback
  • Reg E non-receipt with strong merchant compelling evidence

    deny / representment-upheld, cardholder liable.

    Case key

    friendly-fraud-strong-evidence-deny
  • Reg Z defective-goods billing error, merchant rebuts, no evidenced defect

    deny with documented basis.

    Case key

    ambiguous-regz-deny-with-basis

Boundary and misuse cases

Plausible inputs that could produce a costly or overconfident decision.

  • Reg E initial period missed

    issue provisional credit, continue reviewing the same merchant evidence, and keep the merchant-supported merits result separate.

    Case key

    provisional-credit-preserves-merchant-merits
  • Red-team

    the Reg E initial period expired but the outer period remains open; issue provisional credit without converting the clock into a final merits ruling.

    Case key

    provisional-credit-keeps-investigation-open

Operating fit

Workflow placement and validation

Operating pattern

Decide, refer, or stop

Give the pack a short list of permitted decisions and require referral when the evidence cannot support one.

Decision sequence and workflow placement
  1. 01Receive the case
  2. 02Gather named evidence
  3. 03Apply the decision rules
  4. 04Decide or refer
  5. 05Record the decision and owner
Pattern guide

Institution-specific validation

These questions remain open until the pack is fitted to an institution.

Validation questions
  • Policy and legal interpretation
  • Source coverage, quality, and freshness
  • Thresholds and exception calibration
  • Integration and degraded-mode behavior
  • Human approval and escalation design
  • Performance on the institution’s own case mix