Skip to content
Back to skills

Fraud Pattern Analyst

ASecurity

Identify, test, and prioritise fraud hypotheses and control failures in transactional records. Use for payments, procurement, expenses, payroll, revenue, refunds, vendors, customers, journals, approvals, or access logs.

  • 2 stars
  • 0 votes
  • 0 copies
  • 4 views
  • Added September 4, 2026
ai-agentsrailsgit

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 4, 2026

npx -y skills add Cancellationperiplocagraeca503/legal-ai-skills --skill fraud-pattern-analyst --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Fraud Pattern Analyst?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Fraud Pattern Analyst
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/cancellationperiplocagraeca503-fraud-pattern-analyst/badge)](https://www.skillsdirectory.com/skills/cancellationperiplocagraeca503-fraud-pattern-analyst)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: fraud-pattern-analyst
description: >-
  Identify, test, and prioritise fraud hypotheses and control failures in
  transactional records. Use for payments, procurement, expenses, payroll,
  revenue, refunds, vendors, customers, journals, approvals, or access logs.
---

# Fraud Pattern Analyst

Treat a red flag as a lead, not a finding. Develop plausible fraud and non-fraud
explanations, then test both against preserved source evidence.

## Inputs

Obtain the allegation, objective, period, entities, data dictionary, native
exports, ledger and bank records, master data, contracts, invoices, approvals,
access logs, relationships, and control design. Record missing data and filters.

## Analysis method

1. Preserve raw data and create a repeatable working dataset.
2. Validate meaning, uniqueness, completeness, formats, currencies, signs,
   duplicates, and joins; reconcile control totals where possible.
3. Form competing hypotheses, including error, timing, exception, system
   behaviour, legitimate concentration, and deliberate misconduct.
4. Test relevant indicators: duplicates, round amounts, threshold splitting,
   off-hours activity, sequential invoices, pass-through, overrides, and shared
   addresses, bank details, devices, identifiers, or approvers.
5. Compare suitable peers, cohorts, seasons, locations, and periods.
6. Build relationship links, distinguishing confirmed identity from fuzzy,
   shared, historical, or coincidental matches.
7. Analyse sequences around onboarding, master-data changes, approval, payment,
   refund, reversal, write-off, and access.
8. Trace prioritised exceptions to source documents, system logs, and interviews.
9. Quantify exposure as sourced scenarios without false precision.
10. Map each pattern to expected controls and test design, execution, override,
    and monitoring failures.
11. Rank next steps by evidential value, urgency, preservation risk, cost, and
    risk of alerting subjects.

## Output

Provide a data-quality note, hypothesis matrix, indicator table with innocent
alternatives, linked-party analysis, sample schedule, quantified scenarios,
control-failure analysis, and investigation priorities.

## Guardrails

Do not present suspicion, a score, or a network link as proof. Preserve
exculpatory evidence and apply tests consistently. Do not profile protected
classes, access unauthorised personal data, manipulate records, conceal methods,
or help anyone evade detection.

Files in this skill

  • SKILL.md2.4 KB
  • agents/openai.yaml260 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.